{"ep": "f1-00021", "task": "L2_008", "split": "calib", "reader": "claude-opus-5-5", "arm": "fg_target", "label": "fg_target", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_008|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.0716, "status": "ok", "correct": true, "speedup": 1.0716, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 59.9, "calls": 2, "tokens_in": 8216, "tokens_out": 4018, "ctx_chars": 3639, "ctx_sha": "fc6e42d43dfb4e75", "late_delivered": null, "t_end": 1791552259.3957357, "backend": "triton", "own_frac": 0.022, "ref_ms": 12.71349, "cand_ms": 11.86434, "endpoint": "endpoint-09", "r_wall_s": 12.1, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 2914, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00006", "task": "L2_011", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "show_all", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_011|claude-opus-5-5|1", "design": "f1", "success": true, "score": 4.4037, "status": "ok", "correct": true, "speedup": 4.4037, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 89.4, "calls": 2, "tokens_in": 17802, "tokens_out": 7082, "ctx_chars": 12352, "ctx_sha": "5533cb2a86547de2", "late_delivered": null, "t_end": 1791552288.944642, "backend": "triton", "own_frac": 0.0927, "ref_ms": 27.89699, "cand_ms": 6.33485, "endpoint": "endpoint-10", "r_wall_s": 18.41, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 4702, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00013", "task": "L1_083", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_approach", "label": "no_approach", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L1_083|claude-opus-5-5|1", "design": "f1", "success": true, "score": 3.1114, "status": "ok", "correct": true, "speedup": 3.1114, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 107.5, "calls": 2, "tokens_in": 7990, "tokens_out": 3603, "ctx_chars": 3080, "ctx_sha": "7364d33fba159fed", "late_delivered": null, "t_end": 1791552307.0547667, "backend": "triton", "own_frac": 1.0, "ref_ms": 2.92935, "cand_ms": 0.94148, "endpoint": "endpoint-16", "r_wall_s": 59.49, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 2256, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00008", "task": "L2_057", "split": "calib", "reader": "claude-opus-5-5", "arm": "frontier_gated", "label": "frontier_gated", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_057|claude-opus-5-5|1", "design": "f1", "success": true, "score": 2.172, "status": "ok", "correct": true, "speedup": 2.172, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 109.2, "calls": 2, "tokens_in": 6790, "tokens_out": 1822, "ctx_chars": 3639, "ctx_sha": "fc6e42d43dfb4e75", "late_delivered": null, "t_end": 1791552308.694142, "backend": "triton", "own_frac": 0.2367, "ref_ms": 6.2982, "cand_ms": 2.89978, "endpoint": "endpoint-25", "r_wall_s": 71.38, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 1311, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00015", "task": "L2_035", "split": "calib", "reader": "claude-opus-5-5", "arm": "reader_relative", "label": "reader_relative", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_035|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.2763, "status": "ok", "correct": true, "speedup": 1.2763, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 116.5, "calls": 2, "tokens_in": 13522, "tokens_out": 2025, "ctx_chars": 10461, "ctx_sha": "a45b5774c7161771", "late_delivered": null, "t_end": 1791552316.003643, "backend": "triton", "own_frac": 0.0611, "ref_ms": 17.38101, "cand_ms": 13.61786, "endpoint": "endpoint-17", "r_wall_s": 82.28, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 1130, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00022", "task": "L2_035", "split": "calib", "reader": "claude-opus-5-5", "arm": "fg_target", "label": "fg_target", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_035|claude-opus-5-5|1", "design": "f1", "success": true, "score": 1.2734, "status": "ok", "correct": true, "speedup": 1.2734, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 129.9, "calls": 2, "tokens_in": 7620, "tokens_out": 2944, "ctx_chars": 3639, "ctx_sha": "fc6e42d43dfb4e75", "late_delivered": null, "t_end": 1791552329.477719, "backend": "triton", "own_frac": 0.0629, "ref_ms": 17.44568, "cand_ms": 13.70012, "endpoint": "endpoint-23", "r_wall_s": 61.81, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 2015, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00018", "task": "L1_085", "split": "calib", "reader": "claude-opus-5-5", "arm": "fg_target", "label": "fg_target", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L1_085|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.3145, "status": "ok", "correct": true, "speedup": 1.3145, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 134.4, "calls": 3, "tokens_in": 21274, "tokens_out": 5741, "ctx_chars": 6597, "ctx_sha": "661091d9e159f800", "late_delivered": null, "t_end": 1791552333.8981552, "backend": "triton", "own_frac": 1.0, "ref_ms": 3.73514, "cand_ms": 2.84148, "endpoint": "endpoint-05", "r_wall_s": 60.18, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "triton", "tok_out_first": 2229, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "runtime_error"} {"ep": "f1-00004", "task": "L2_071", "split": "calib", "reader": "claude-opus-5-5", "arm": "fg_target", "label": "fg_target", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_071|claude-opus-5-5|1", "design": "f1", "success": true, "score": 1.2723, "status": "ok", "correct": true, "speedup": 1.2723, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 135.1, "calls": 2, "tokens_in": 7648, "tokens_out": 3712, "ctx_chars": 3639, "ctx_sha": "fc6e42d43dfb4e75", "late_delivered": null, "t_end": 1791552334.5807142, "backend": "triton", "own_frac": 1.0, "ref_ms": 3.66615, "cand_ms": 2.88149, "endpoint": "endpoint-23", "r_wall_s": 71.54, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 2567, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00023", "task": "L1_065", "split": "calib", "reader": "claude-opus-5-5", "arm": "frontier_gated", "label": "frontier_gated", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L1_065|claude-opus-5-5|1", "design": "f1", "success": false, "score": 0.0, "status": "runtime_error", "correct": false, "speedup": 0.0, "infra": false, "infra_reruns": 0, "error": "OutOfMemoryError: CUDA out of memory. Tried to allocate 1024.00 MiB. GPU 0 has a total capacity of 39.49 GiB of which 98.81 MiB is free. Including non-PyTorch memory, this process has 1.16 GiB memory in use. Of the allocated memory 768.33 MiB is allocated by PyTorch, and 1.67 MiB is reserved by PyTo", "wall_s": 144.1, "calls": 2, "tokens_in": 9443, "tokens_out": 8615, "ctx_chars": 3256, "ctx_sha": "8fa366c59fab0c7f", "late_delivered": null, "t_end": 1791552343.597087, "backend": "triton", "own_frac": null, "ref_ms": null, "cand_ms": null, "endpoint": "endpoint-12", "r_wall_s": 6.8, "flags": null, "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 6290, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00007", "task": "L2_027", "split": "calib", "reader": "claude-opus-5-5", "arm": "frontier_gated", "label": "frontier_gated", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_027|claude-opus-5-5|1", "design": "f1", "success": true, "score": 3.9752, "status": "ok", "correct": true, "speedup": 3.9752, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 144.1, "calls": 2, "tokens_in": 13805, "tokens_out": 11872, "ctx_chars": 7376, "ctx_sha": "cc69cebe7b058084", "late_delivered": null, "t_end": 1791552343.6699357, "backend": "cuda", "own_frac": 0.9996, "ref_ms": 20.06029, "cand_ms": 5.04642, "endpoint": "endpoint-10", "r_wall_s": 18.84, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 8653, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00017", "task": "L3_004", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_approach", "label": "no_approach", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L3_004|claude-opus-5-5|1", "design": "f1", "success": true, "score": 4.6166, "status": "ok", "correct": true, "speedup": 4.6166, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 156.7, "calls": 2, "tokens_in": 9386, "tokens_out": 7636, "ctx_chars": 2932, "ctx_sha": "e5406f0f931ba9ce", "late_delivered": null, "t_end": 1791552356.2403111, "backend": "cuda", "own_frac": 0.7768, "ref_ms": 2.30836, "cand_ms": 0.50001, "endpoint": "endpoint-08", "r_wall_s": 48.49, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 5005, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00009", "task": "L2_024", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_approach", "label": "no_approach", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_024|claude-opus-5-5|1", "design": "f1", "success": true, "score": 2.2573, "status": "ok", "correct": true, "speedup": 2.2573, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 165.2, "calls": 3, "tokens_in": 17920, "tokens_out": 10753, "ctx_chars": 3080, "ctx_sha": "7364d33fba159fed", "late_delivered": null, "t_end": 1791552364.6973891, "backend": "cuda", "own_frac": 1.0, "ref_ms": 6.5848, "cand_ms": 2.91708, "endpoint": "endpoint-04", "r_wall_s": 43.92, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 5177, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00024", "task": "L2_018", "split": "calib", "reader": "claude-opus-5-5", "arm": "lessons_only", "label": "lessons_only", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_018|claude-opus-5-5|1", "design": "f1", "success": true, "score": 1.0098, "status": "ok", "correct": true, "speedup": 1.0098, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 107.6, "calls": 3, "tokens_in": 12431, "tokens_out": 7355, "ctx_chars": 1982, "ctx_sha": "54e16caf9bef9eec", "late_delivered": null, "t_end": 1791552366.946343, "backend": "triton", "own_frac": 0.0028, "ref_ms": 8.01735, "cand_ms": 7.9398, "endpoint": "endpoint-17", "r_wall_s": 13.48, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "triton", "tok_out_first": 2827, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "incorrect"} {"ep": "f1-00011", "task": "L1_072", "split": "calib", "reader": "claude-opus-5-5", "arm": "lessons_only", "label": "lessons_only", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L1_072|claude-opus-5-5|1", "design": "f1", "success": true, "score": 5.4624, "status": "ok", "correct": true, "speedup": 5.4624, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 174.5, "calls": 2, "tokens_in": 10628, "tokens_out": 14909, "ctx_chars": 1982, "ctx_sha": "54e16caf9bef9eec", "late_delivered": null, "t_end": 1791552374.0636644, "backend": "cuda", "own_frac": 1.0, "ref_ms": 4.3869, "cand_ms": 0.8031, "endpoint": "endpoint-07", "r_wall_s": 29.89, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 10759, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00010", "task": "L1_077", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_approach", "label": "no_approach", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L1_077|claude-opus-5-5|1", "design": "f1", "success": true, "score": 1.5855, "status": "ok", "correct": true, "speedup": 1.5855, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 180.3, "calls": 2, "tokens_in": 9208, "tokens_out": 10361, "ctx_chars": 3030, "ctx_sha": "e6667ea63ccdb086", "late_delivered": null, "t_end": 1791552379.835807, "backend": "triton", "own_frac": 1.0, "ref_ms": 8.6115, "cand_ms": 5.43143, "endpoint": "endpoint-26", "r_wall_s": 60.25, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 8168, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00012", "task": "L2_027", "split": "calib", "reader": "claude-opus-5-5", "arm": "reader_relative", "label": "reader_relative", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_027|claude-opus-5-5|0", "design": "f1", "success": true, "score": 4.0309, "status": "ok", "correct": true, "speedup": 4.0309, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 187.0, "calls": 2, "tokens_in": 18069, "tokens_out": 13859, "ctx_chars": 12166, "ctx_sha": "9848868d582a6045", "late_delivered": null, "t_end": 1791552386.5311558, "backend": "cuda", "own_frac": 0.9996, "ref_ms": 20.11184, "cand_ms": 4.98938, "endpoint": "endpoint-16", "r_wall_s": 20.93, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 10450, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00000", "task": "L2_074", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_approach", "label": "no_approach", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_074|claude-opus-5-5|0", "design": "f1", "success": true, "score": 9.3588, "status": "ok", "correct": true, "speedup": 9.3588, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 188.4, "calls": 3, "tokens_in": 18653, "tokens_out": 13997, "ctx_chars": 3030, "ctx_sha": "e6667ea63ccdb086", "late_delivered": null, "t_end": 1791552387.977667, "backend": "cuda", "own_frac": 1.0, "ref_ms": 5.95248, "cand_ms": 0.63603, "endpoint": "endpoint-08", "r_wall_s": 23.31, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 8206, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00019", "task": "L2_082", "split": "calib", "reader": "claude-opus-5-5", "arm": "none", "label": "none", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_082|claude-opus-5-5|0", "design": "f1", "success": true, "score": 3.9567, "status": "ok", "correct": true, "speedup": 3.9567, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 192.2, "calls": 3, "tokens_in": 14685, "tokens_out": 11602, "ctx_chars": 0, "ctx_sha": "e3b0c44298fc1c14", "late_delivered": null, "t_end": 1791552391.7299757, "backend": "cuda", "own_frac": 0.9994, "ref_ms": 19.46013, "cand_ms": 4.91833, "endpoint": "endpoint-14", "r_wall_s": 33.43, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 6099, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00002", "task": "L1_072", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_approach", "label": "no_approach", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L1_072|claude-opus-5-5|0", "design": "f1", "success": true, "score": 5.1661, "status": "ok", "correct": true, "speedup": 5.1661, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 198.7, "calls": 2, "tokens_in": 11058, "tokens_out": 15151, "ctx_chars": 3030, "ctx_sha": "e6667ea63ccdb086", "late_delivered": null, "t_end": 1791552398.2530046, "backend": "cuda", "own_frac": 1.0, "ref_ms": 4.39376, "cand_ms": 0.8505, "endpoint": "endpoint-12", "r_wall_s": 28.16, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 11375, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00028", "task": "L3_004", "split": "calib", "reader": "claude-opus-5-5", "arm": "frontier_gated", "label": "frontier_gated", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L3_004|claude-opus-5-5|1", "design": "f1", "success": true, "score": 4.4009, "status": "ok", "correct": true, "speedup": 4.4009, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 91.6, "calls": 2, "tokens_in": 8861, "tokens_out": 6992, "ctx_chars": 2759, "ctx_sha": "e8b611ac667b08e3", "late_delivered": null, "t_end": 1791552407.6208162, "backend": "cuda", "own_frac": 0.7862, "ref_ms": 2.30211, "cand_ms": 0.5231, "endpoint": "endpoint-26", "r_wall_s": 18.75, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 4543, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00020", "task": "L1_071", "split": "calib", "reader": "claude-opus-5-5", "arm": "frontier_gated", "label": "frontier_gated", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L1_071|claude-opus-5-5|1", "design": "f1", "success": false, "score": 0.8412, "status": "ok", "correct": true, "speedup": 0.8412, "infra": false, "infra_reruns": 0, "error": "ok", "wall_s": 211.0, "calls": 2, "tokens_in": 8883, "tokens_out": 6082, "ctx_chars": 3256, "ctx_sha": "8fa366c59fab0c7f", "late_delivered": null, "t_end": 1791552410.5842252, "backend": "triton", "own_frac": 0.999, "ref_ms": 3.3225, "cand_ms": 3.94983, "endpoint": "endpoint-19", "r_wall_s": 83.28, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 4270, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00003", "task": "L1_080", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "show_all", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L1_080|claude-opus-5-5|1", "design": "f1", "success": false, "score": 0.8774, "status": "ok", "correct": true, "speedup": 0.8774, "infra": false, "infra_reruns": 0, "error": "ok", "wall_s": 215.5, "calls": 3, "tokens_in": 27304, "tokens_out": 8583, "ctx_chars": 10864, "ctx_sha": "0b6222244e19e377", "late_delivered": null, "t_end": 1791552415.040625, "backend": "triton", "own_frac": 0.9998, "ref_ms": 20.17069, "cand_ms": 22.98926, "endpoint": "endpoint-25", "r_wall_s": 59.27, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "triton", "tok_out_first": 3997, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "runtime_error"} {"ep": "f1-00029", "task": "L3_046", "split": "calib", "reader": "claude-opus-5-5", "arm": "lessons_only", "label": "lessons_only", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L3_046|claude-opus-5-5|1", "design": "f1", "success": false, "score": 0.0, "status": "runtime_error", "correct": false, "speedup": 0.0, "infra": false, "infra_reruns": 0, "error": "OutOfMemoryError: CUDA out of memory. Tried to allocate 800.00 MiB. GPU 0 has a total capacity of 39.49 GiB of which 284.81 MiB is free. Including non-PyTorch memory, this process has 1.05 GiB memory in use. Of the allocated memory 656.16 MiB is allocated by PyTorch, and 1.84 MiB is reserved by PyTo", "wall_s": 98.1, "calls": 2, "tokens_in": 9585, "tokens_out": 7600, "ctx_chars": 1982, "ctx_sha": "54e16caf9bef9eec", "late_delivered": null, "t_end": 1791552427.5407495, "backend": "triton", "own_frac": null, "ref_ms": null, "cand_ms": null, "endpoint": "endpoint-12", "r_wall_s": 5.14, "flags": null, "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 5545, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00026", "task": "L1_050", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_approach", "label": "no_approach", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L1_050|claude-opus-5-5|0", "design": "f1", "success": false, "score": 0.6426, "status": "ok", "correct": true, "speedup": 0.6426, "infra": false, "infra_reruns": 0, "error": "ok", "wall_s": 125.4, "calls": 2, "tokens_in": 7337, "tokens_out": 5373, "ctx_chars": 3080, "ctx_sha": "7364d33fba159fed", "late_delivered": null, "t_end": 1791552432.4640973, "backend": "triton", "own_frac": 1.0, "ref_ms": 5.33799, "cand_ms": 8.3067, "endpoint": "endpoint-07", "r_wall_s": 51.47, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 3786, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00001", "task": "L3_039", "split": "calib", "reader": "claude-opus-5-5", "arm": "fg_target", "label": "fg_target", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L3_039|claude-opus-5-5|1", "design": "f1", "success": true, "score": 4.202, "status": "ok", "correct": true, "speedup": 4.202, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 236.0, "calls": 3, "tokens_in": 19973, "tokens_out": 19561, "ctx_chars": 2260, "ctx_sha": "9b6af621f8f8a88b", "late_delivered": null, "t_end": 1791552435.5386245, "backend": "cuda", "own_frac": 0.982, "ref_ms": 32.17269, "cand_ms": 7.65656, "endpoint": "endpoint-01", "r_wall_s": 19.44, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 12269, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00016", "task": "L3_049", "split": "calib", "reader": "claude-opus-5-5", "arm": "reader_relative", "label": "reader_relative", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L3_049|claude-opus-5-5|1", "design": "f1", "success": true, "score": 10.0, "status": "ok", "correct": true, "speedup": 11.6036, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 236.6, "calls": 5, "tokens_in": 32360, "tokens_out": 15408, "ctx_chars": 2902, "ctx_sha": "98e2cfd37ea90942", "late_delivered": null, "t_end": 1791552436.098298, "backend": "triton", "own_frac": 0.7009, "ref_ms": 11.96737, "cand_ms": 1.03135, "endpoint": "endpoint-19", "r_wall_s": 18.8, "flags": ["excessive_speedup>10.0x: audit"], "submit_turn": 4, "turns": 5, "tech_src": "triton", "tok_out_first": 3617, "stops": ["end_turn", null, "end_turn", "end_turn", "end_turn"], "first_check": "incorrect"} {"ep": "f1-00033", "task": "L2_073", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_numbers", "label": "no_numbers", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_073|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.4638, "status": "ok", "correct": true, "speedup": 1.4638, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 108.6, "calls": 2, "tokens_in": 13440, "tokens_out": 4415, "ctx_chars": 9582, "ctx_sha": "10cc9f53178dc774", "late_delivered": null, "t_end": 1791552452.2361162, "backend": "triton", "own_frac": 0.3811, "ref_ms": 5.61634, "cand_ms": 3.8369, "endpoint": "endpoint-14", "r_wall_s": 53.22, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 2911, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00005", "task": "L2_082", "split": "calib", "reader": "claude-opus-5-5", "arm": "fg_target", "label": "fg_target", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_082|claude-opus-5-5|1", "design": "f1", "success": false, "score": 0.0, "status": "runtime_error", "correct": false, "speedup": 0.0, "infra": false, "infra_reruns": 0, "error": "AcceleratorError: CUDA error: out of memory\nSearch for `cudaErrorMemoryAllocation' in https://docs.nvidia.com/cuda/cuda-runtime-api/group__CUDART__TYPES.html for more information.\nCUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.\n", "wall_s": 281.3, "calls": 3, "tokens_in": 19096, "tokens_out": 15820, "ctx_chars": 3639, "ctx_sha": "fc6e42d43dfb4e75", "late_delivered": null, "t_end": 1791552480.8243833, "backend": "cuda", "own_frac": null, "ref_ms": null, "cand_ms": null, "endpoint": "endpoint-08", "r_wall_s": 45.76, "flags": null, "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 9733, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00039", "task": "L2_071", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "show_all", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_071|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.235, "status": "ok", "correct": true, "speedup": 1.235, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 100.1, "calls": 2, "tokens_in": 13810, "tokens_out": 2124, "ctx_chars": 11132, "ctx_sha": "4caa9b6e3a0949e1", "late_delivered": null, "t_end": 1791552486.6469245, "backend": "triton", "own_frac": 0.236, "ref_ms": 3.6442, "cand_ms": 2.95072, "endpoint": "endpoint-16", "r_wall_s": 57.98, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 1496, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00031", "task": "L2_018", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_approach", "label": "no_approach", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_018|claude-opus-5-5|1", "design": "f1", "success": true, "score": 1.009, "status": "ok", "correct": true, "speedup": 1.009, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 159.1, "calls": 4, "tokens_in": 23066, "tokens_out": 8281, "ctx_chars": 3288, "ctx_sha": "dc1a3e459ffe3f99", "late_delivered": null, "t_end": 1791552493.7035038, "backend": "triton", "own_frac": 0.003, "ref_ms": 8.00811, "cand_ms": 7.93642, "endpoint": "endpoint-12", "r_wall_s": 13.53, "flags": [], "submit_turn": 3, "turns": 4, "tech_src": "triton", "tok_out_first": 2968, "stops": ["end_turn", "end_turn", "end_turn", "end_turn"], "first_check": "incorrect"} {"ep": "f1-00035", "task": "L3_008", "split": "calib", "reader": "claude-opus-5-5", "arm": "fg_target", "label": "fg_target", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L3_008|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.2448, "status": "ok", "correct": true, "speedup": 1.2448, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 137.3, "calls": 2, "tokens_in": 9262, "tokens_out": 9480, "ctx_chars": 2759, "ctx_sha": "e8b611ac667b08e3", "late_delivered": null, "t_end": 1791552501.9623747, "backend": "triton", "own_frac": 0.0913, "ref_ms": 5.25061, "cand_ms": 4.21799, "endpoint": "endpoint-24", "r_wall_s": 26.08, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 6570, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00034", "task": "L3_008", "split": "calib", "reader": "claude-opus-5-5", "arm": "frontier_gated", "label": "frontier_gated", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L3_008|claude-opus-5-5|1", "design": "f1", "success": true, "score": 1.1813, "status": "ok", "correct": true, "speedup": 1.1813, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 152.4, "calls": 2, "tokens_in": 9082, "tokens_out": 10334, "ctx_chars": 2759, "ctx_sha": "e8b611ac667b08e3", "late_delivered": null, "t_end": 1791552508.6641476, "backend": "triton", "own_frac": 0.0801, "ref_ms": 5.27887, "cand_ms": 4.46874, "endpoint": "endpoint-19", "r_wall_s": 32.65, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 7669, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00038", "task": "L2_079", "split": "calib", "reader": "claude-opus-5-5", "arm": "none", "label": "none", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_079|claude-opus-5-5|0", "design": "f1", "success": true, "score": 2.6864, "status": "ok", "correct": true, "speedup": 2.6864, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 131.8, "calls": 2, "tokens_in": 6890, "tokens_out": 7692, "ctx_chars": 0, "ctx_sha": "e3b0c44298fc1c14", "late_delivered": null, "t_end": 1791552511.6724815, "backend": "cuda", "own_frac": 0.9972, "ref_ms": 2.18079, "cand_ms": 0.81179, "endpoint": "endpoint-10", "r_wall_s": 18.4, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 4867, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00027", "task": "L1_080", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_approach", "label": "no_approach", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L1_080|claude-opus-5-5|1", "design": "f1", "success": false, "score": 0.9407, "status": "ok", "correct": true, "speedup": 0.9407, "infra": false, "infra_reruns": 0, "error": "ok", "wall_s": 203.1, "calls": 2, "tokens_in": 10264, "tokens_out": 13924, "ctx_chars": 3080, "ctx_sha": "7364d33fba159fed", "late_delivered": null, "t_end": 1791552511.803504, "backend": "cuda", "own_frac": 0.9998, "ref_ms": 20.12284, "cand_ms": 21.39107, "endpoint": "endpoint-09", "r_wall_s": 62.89, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 10589, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00030", "task": "L2_018", "split": "calib", "reader": "claude-opus-5-5", "arm": "none", "label": "none", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_018|claude-opus-5-5|0", "design": "f1", "success": true, "score": 2.9944, "status": "ok", "correct": true, "speedup": 2.9944, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 178.6, "calls": 3, "tokens_in": 13205, "tokens_out": 13195, "ctx_chars": 0, "ctx_sha": "e3b0c44298fc1c14", "late_delivered": null, "t_end": 1791552512.5024576, "backend": "cuda", "own_frac": 0.0815, "ref_ms": 8.01078, "cand_ms": 2.67528, "endpoint": "endpoint-17", "r_wall_s": 21.73, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 3227, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "incorrect"} {"ep": "f1-00043", "task": "L1_085", "split": "calib", "reader": "claude-opus-5-5", "arm": "reader_relative", "label": "reader_relative", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L1_085|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.3106, "status": "ok", "correct": true, "speedup": 1.3106, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 120.2, "calls": 2, "tokens_in": 16318, "tokens_out": 3784, "ctx_chars": 11490, "ctx_sha": "2e797d2c54777bc5", "late_delivered": null, "t_end": 1791552527.8383162, "backend": "triton", "own_frac": 1.0, "ref_ms": 3.73749, "cand_ms": 2.85183, "endpoint": "endpoint-23", "r_wall_s": 54.46, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 2190, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00040", "task": "L2_024", "split": "calib", "reader": "claude-opus-5-5", "arm": "lessons_only", "label": "lessons_only", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_024|claude-opus-5-5|1", "design": "f1", "success": false, "score": 0.2506, "status": "ok", "correct": true, "speedup": 0.2506, "infra": false, "infra_reruns": 0, "error": "ok", "wall_s": 140.4, "calls": 3, "tokens_in": 13963, "tokens_out": 8012, "ctx_chars": 1982, "ctx_sha": "54e16caf9bef9eec", "late_delivered": null, "t_end": 1791552528.3762772, "backend": "cuda", "own_frac": 0.9997, "ref_ms": 2.37633, "cand_ms": 9.48336, "endpoint": "endpoint-26", "r_wall_s": 21.19, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 4183, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00044", "task": "L2_073", "split": "calib", "reader": "claude-opus-5-5", "arm": "frontier_gated", "label": "frontier_gated", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_073|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.6237, "status": "ok", "correct": true, "speedup": 1.6237, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 128.6, "calls": 2, "tokens_in": 7926, "tokens_out": 4677, "ctx_chars": 3639, "ctx_sha": "fc6e42d43dfb4e75", "late_delivered": null, "t_end": 1791552539.1773334, "backend": "triton", "own_frac": 0.3231, "ref_ms": 5.61301, "cand_ms": 3.45699, "endpoint": "endpoint-14", "r_wall_s": 66.97, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 3221, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00049", "task": "L2_027", "split": "calib", "reader": "claude-opus-5-5", "arm": "none", "label": "none", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_027|claude-opus-5-5|0", "design": "f1", "success": false, "score": 0.375, "status": "ok", "correct": true, "speedup": 0.375, "infra": false, "infra_reruns": 0, "error": "ok", "wall_s": 131.5, "calls": 2, "tokens_in": 7015, "tokens_out": 10421, "ctx_chars": 0, "ctx_sha": "e3b0c44298fc1c14", "late_delivered": null, "t_end": 1791552567.5585127, "backend": "cuda", "own_frac": 0.0001, "ref_ms": 20.0681, "cand_ms": 53.51635, "endpoint": "endpoint-09", "r_wall_s": 21.41, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 7469, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00041", "task": "L1_080", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_approach", "label": "no_approach", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L1_080|claude-opus-5-5|0", "design": "f1", "success": false, "score": 0.7573, "status": "ok", "correct": true, "speedup": 0.7573, "infra": false, "infra_reruns": 0, "error": "ok", "wall_s": 181.1, "calls": 2, "tokens_in": 8942, "tokens_out": 6259, "ctx_chars": 3080, "ctx_sha": "7364d33fba159fed", "late_delivered": null, "t_end": 1791552572.8138947, "backend": "triton", "own_frac": 0.9998, "ref_ms": 20.15357, "cand_ms": 26.61235, "endpoint": "endpoint-18", "r_wall_s": 79.99, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 4287, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00014", "task": "L3_039", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_approach", "label": "no_approach", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L3_039|claude-opus-5-5|1", "design": "f1", "success": true, "score": 8.1766, "status": "ok", "correct": true, "speedup": 8.1766, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 373.5, "calls": 4, "tokens_in": 46302, "tokens_out": 33916, "ctx_chars": 3355, "ctx_sha": "cc270772e526c41f", "late_delivered": null, "t_end": 1791552572.99, "backend": "cuda", "own_frac": 0.9995, "ref_ms": 31.28288, "cand_ms": 3.82588, "endpoint": "endpoint-24", "r_wall_s": 19.85, "flags": [], "submit_turn": 3, "turns": 4, "tech_src": "cuda", "tok_out_first": 18711, "stops": ["end_turn", "end_turn", "end_turn", "end_turn"], "first_check": "compile_error"} {"ep": "f1-00025", "task": "L3_049", "split": "calib", "reader": "claude-opus-5-5", "arm": "frontier_gated", "label": "frontier_gated", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L3_049|claude-opus-5-5|1", "design": "f1", "success": true, "score": 10.0, "status": "ok", "correct": true, "speedup": 12.3482, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 292.2, "calls": 4, "tokens_in": 37402, "tokens_out": 19099, "ctx_chars": 2578, "ctx_sha": "7b4237bf96a5394d", "late_delivered": null, "t_end": 1791552581.1346226, "backend": "cuda", "own_frac": 0.9615, "ref_ms": 12.02747, "cand_ms": 0.97403, "endpoint": "endpoint-10", "r_wall_s": 51.06, "flags": ["excessive_speedup>10.0x: audit"], "submit_turn": 3, "turns": 4, "tech_src": "cuda", "tok_out_first": 7778, "stops": ["end_turn", "end_turn", "end_turn", "end_turn"], "first_check": "incorrect"} {"ep": "f1-00047", "task": "L1_077", "split": "calib", "reader": "claude-opus-5-5", "arm": "lessons_only", "label": "lessons_only", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L1_077|claude-opus-5-5|0", "design": "f1", "success": true, "score": 2.8335, "status": "ok", "correct": true, "speedup": 2.8335, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 175.2, "calls": 2, "tokens_in": 7664, "tokens_out": 7928, "ctx_chars": 1982, "ctx_sha": "54e16caf9bef9eec", "late_delivered": null, "t_end": 1791552607.6670353, "backend": "triton", "own_frac": 0.941, "ref_ms": 8.62724, "cand_ms": 3.04473, "endpoint": "endpoint-10", "r_wall_s": 78.55, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 6438, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00059", "task": "L2_046", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "late_show_all", "delivery": "late", "rep": 0, "p_arm": 0.111111, "block": "L2_046|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.5633, "status": "ok", "correct": true, "speedup": 1.5633, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 82.1, "calls": 2, "tokens_in": 9560, "tokens_out": 3561, "ctx_chars": 10210, "ctx_sha": "ee72d31f6aad351d", "late_delivered": true, "t_end": 1791552609.9675515, "backend": "triton", "own_frac": 0.0891, "ref_ms": 19.34576, "cand_ms": 12.37531, "endpoint": "endpoint-14", "r_wall_s": 37.87, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 2384, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00036", "task": "L3_039", "split": "calib", "reader": "claude-opus-5-5", "arm": "none", "label": "none", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L3_039|claude-opus-5-5|1", "design": "f1", "success": true, "score": 6.6543, "status": "ok", "correct": true, "speedup": 6.6543, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 245.8, "calls": 3, "tokens_in": 19495, "tokens_out": 21924, "ctx_chars": 0, "ctx_sha": "e3b0c44298fc1c14", "late_delivered": null, "t_end": 1791552612.7745245, "backend": "cuda", "own_frac": 0.9884, "ref_ms": 31.3488, "cand_ms": 4.71106, "endpoint": "endpoint-11", "r_wall_s": 28.72, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 13326, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00046", "task": "L3_049", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_approach", "label": "no_approach", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L3_049|claude-opus-5-5|1", "design": "f1", "success": true, "score": 3.6212, "status": "ok", "correct": true, "speedup": 3.6212, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 193.0, "calls": 4, "tokens_in": 26537, "tokens_out": 14163, "ctx_chars": 2425, "ctx_sha": "5c5e8811718de973", "late_delivered": null, "t_end": 1791552620.5823777, "backend": "triton", "own_frac": 0.0409, "ref_ms": 11.96118, "cand_ms": 3.30307, "endpoint": "endpoint-05", "r_wall_s": 30.17, "flags": [], "submit_turn": 3, "turns": 4, "tech_src": "triton", "tok_out_first": 3992, "stops": ["end_turn", "end_turn", "end_turn", "end_turn"], "first_check": "incorrect"} {"ep": "f1-00058", "task": "L1_083", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_numbers", "label": "no_numbers", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L1_083|claude-opus-5-5|0", "design": "f1", "success": true, "score": 3.254, "status": "ok", "correct": true, "speedup": 3.254, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 110.9, "calls": 2, "tokens_in": 14465, "tokens_out": 3533, "ctx_chars": 10351, "ctx_sha": "31905f761c946397", "late_delivered": null, "t_end": 1791552623.4514108, "backend": "triton", "own_frac": 1.0, "ref_ms": 2.93366, "cand_ms": 0.90157, "endpoint": "endpoint-07", "r_wall_s": 63.99, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 2233, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00048", "task": "L3_004", "split": "calib", "reader": "claude-opus-5-5", "arm": "reader_relative", "label": "reader_relative", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L3_004|claude-opus-5-5|0", "design": "f1", "success": true, "score": 3.8638, "status": "ok", "correct": true, "speedup": 3.8638, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 196.1, "calls": 3, "tokens_in": 28733, "tokens_out": 18300, "ctx_chars": 7026, "ctx_sha": "150032ec66a5d1b9", "late_delivered": null, "t_end": 1791552631.6742077, "backend": "cuda", "own_frac": 1.0, "ref_ms": 2.37536, "cand_ms": 0.61478, "endpoint": "endpoint-10", "r_wall_s": 19.54, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 8943, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "runtime_error"} {"ep": "f1-00051", "task": "L2_011", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "show_all", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_011|claude-opus-5-5|0", "design": "f1", "success": true, "score": 4.3765, "status": "ok", "correct": true, "speedup": 4.3765, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 152.2, "calls": 2, "tokens_in": 17603, "tokens_out": 7447, "ctx_chars": 12352, "ctx_sha": "5533cb2a86547de2", "late_delivered": null, "t_end": 1791552632.9910295, "backend": "triton", "own_frac": 0.0867, "ref_ms": 27.98288, "cand_ms": 6.39394, "endpoint": "endpoint-09", "r_wall_s": 59.07, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 5179, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00037", "task": "L1_071", "split": "calib", "reader": "claude-opus-5-5", "arm": "frontier_gated", "label": "frontier_gated", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L1_071|claude-opus-5-5|0", "design": "f1", "success": false, "score": 0.0, "status": "runtime_error", "correct": false, "speedup": 0.0, "infra": false, "infra_reruns": 0, "error": "OutOfMemoryError: CUDA out of memory. Tried to allocate 1024.00 MiB. GPU 0 has a total capacity of 39.49 GiB of which 104.81 MiB is free. Including non-PyTorch memory, this process has 1.16 GiB memory in use. Of the allocated memory 768.04 MiB is allocated by PyTorch, and 1.96 MiB is reserved by PyT", "wall_s": 260.9, "calls": 3, "tokens_in": 18980, "tokens_out": 12910, "ctx_chars": 3256, "ctx_sha": "8fa366c59fab0c7f", "late_delivered": null, "t_end": 1791552634.9899948, "backend": "cuda", "own_frac": null, "ref_ms": null, "cand_ms": null, "endpoint": "endpoint-12", "r_wall_s": 44.4, "flags": null, "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 7372, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00060", "task": "L2_079", "split": "calib", "reader": "claude-opus-5-5", "arm": "fg_target", "label": "fg_target", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_079|claude-opus-5-5|1", "design": "f1", "success": false, "score": 0.1379, "status": "ok", "correct": true, "speedup": 0.1379, "infra": false, "infra_reruns": 0, "error": "ok", "wall_s": 113.1, "calls": 2, "tokens_in": 9488, "tokens_out": 6793, "ctx_chars": 3639, "ctx_sha": "fc6e42d43dfb4e75", "late_delivered": null, "t_end": 1791552641.4469569, "backend": "triton", "own_frac": 0.9999, "ref_ms": 2.16101, "cand_ms": 15.67241, "endpoint": "endpoint-23", "r_wall_s": 31.63, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 4307, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00032", "task": "L3_036", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "show_all", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L3_036|claude-opus-5-5|0", "design": "f1", "success": true, "score": 5.2674, "status": "ok", "correct": true, "speedup": 5.2674, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 306.2, "calls": 4, "tokens_in": 46912, "tokens_out": 27946, "ctx_chars": 8224, "ctx_sha": "71fd1a97cbd02e3b", "late_delivered": null, "t_end": 1791552649.759183, "backend": "cuda", "own_frac": 0.9798, "ref_ms": 32.03503, "cand_ms": 6.08173, "endpoint": "endpoint-13", "r_wall_s": 19.84, "flags": [], "submit_turn": 3, "turns": 4, "tech_src": "cuda", "tok_out_first": 15255, "stops": ["end_turn", "end_turn", "end_turn", "end_turn"], "first_check": "runtime_error"} {"ep": "f1-00055", "task": "L2_027", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "late_show_all", "delivery": "late", "rep": 0, "p_arm": 0.111111, "block": "L2_027|claude-opus-5-5|0", "design": "f1", "success": true, "score": 3.4669, "status": "ok", "correct": true, "speedup": 3.4669, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 143.2, "calls": 2, "tokens_in": 11884, "tokens_out": 10471, "ctx_chars": 11535, "ctx_sha": "114ec077101d70e2", "late_delivered": true, "t_end": 1791552651.8844342, "backend": "cuda", "own_frac": 0.0006, "ref_ms": 20.09196, "cand_ms": 5.79538, "endpoint": "endpoint-08", "r_wall_s": 20.64, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 7600, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00053", "task": "L2_074", "split": "calib", "reader": "claude-opus-5-5", "arm": "lessons_only", "label": "lessons_only", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_074|claude-opus-5-5|1", "design": "f1", "success": true, "score": 8.2372, "status": "ok", "correct": true, "speedup": 8.2372, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 161.2, "calls": 3, "tokens_in": 16759, "tokens_out": 11568, "ctx_chars": 1982, "ctx_sha": "54e16caf9bef9eec", "late_delivered": null, "t_end": 1791552654.8628168, "backend": "cuda", "own_frac": 1.0, "ref_ms": 5.99321, "cand_ms": 0.72758, "endpoint": "endpoint-07", "r_wall_s": 30.09, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 6222, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00052", "task": "L2_018", "split": "calib", "reader": "claude-opus-5-5", "arm": "fg_target", "label": "fg_target", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_018|claude-opus-5-5|1", "design": "f1", "success": true, "score": 3.3608, "status": "ok", "correct": true, "speedup": 3.3608, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 171.6, "calls": 4, "tokens_in": 22950, "tokens_out": 12916, "ctx_chars": 2990, "ctx_sha": "2d414f7ee5025848", "late_delivered": null, "t_end": 1791552658.2900124, "backend": "triton", "own_frac": 0.1154, "ref_ms": 8.0082, "cand_ms": 2.38284, "endpoint": "endpoint-01", "r_wall_s": 21.53, "flags": [], "submit_turn": 3, "turns": 4, "tech_src": "triton", "tok_out_first": 2599, "stops": ["end_turn", "end_turn", "end_turn", "end_turn"], "first_check": "incorrect"} {"ep": "f1-00050", "task": "L1_078", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "show_all", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L1_078|claude-opus-5-5|1", "design": "f1", "success": true, "score": 1.1991, "status": "ok", "correct": true, "speedup": 1.1991, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 212.8, "calls": 3, "tokens_in": 28957, "tokens_out": 8888, "ctx_chars": 11582, "ctx_sha": "a74f6646d7158b72", "late_delivered": null, "t_end": 1791552665.0779185, "backend": "triton", "own_frac": 1.0, "ref_ms": 35.64503, "cand_ms": 29.72567, "endpoint": "endpoint-26", "r_wall_s": 77.68, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "triton", "tok_out_first": 4460, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "runtime_error"} {"ep": "f1-00061", "task": "L2_018", "split": "calib", "reader": "claude-opus-5-5", "arm": "frontier_gated", "label": "frontier_gated", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_018|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.0104, "status": "ok", "correct": true, "speedup": 1.0104, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 128.9, "calls": 4, "tokens_in": 21108, "tokens_out": 7458, "ctx_chars": 2990, "ctx_sha": "2d414f7ee5025848", "late_delivered": null, "t_end": 1791552668.1205077, "backend": "triton", "own_frac": 0.0028, "ref_ms": 8.01165, "cand_ms": 7.92944, "endpoint": "endpoint-17", "r_wall_s": 13.62, "flags": [], "submit_turn": 3, "turns": 4, "tech_src": "triton", "tok_out_first": 2447, "stops": ["end_turn", "end_turn", "end_turn", "end_turn"], "first_check": "incorrect"} {"ep": "f1-00045", "task": "L1_072", "split": "calib", "reader": "claude-opus-5-5", "arm": "reader_relative", "label": "reader_relative", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L1_072|claude-opus-5-5|1", "design": "f1", "success": true, "score": 4.8627, "status": "ok", "correct": true, "speedup": 4.8627, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 257.3, "calls": 3, "tokens_in": 35297, "tokens_out": 20864, "ctx_chars": 11807, "ctx_sha": "04bab9b1ff7e3fde", "late_delivered": null, "t_end": 1791552672.342327, "backend": "cuda", "own_frac": 1.0, "ref_ms": 4.38426, "cand_ms": 0.90161, "endpoint": "endpoint-09", "r_wall_s": 37.18, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 12506, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00056", "task": "L1_085", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_approach", "label": "no_approach", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L1_085|claude-opus-5-5|1", "design": "f1", "success": true, "score": 2.1809, "status": "ok", "correct": true, "speedup": 2.1809, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 188.8, "calls": 2, "tokens_in": 10838, "tokens_out": 10429, "ctx_chars": 3080, "ctx_sha": "7364d33fba159fed", "late_delivered": null, "t_end": 1791552700.5036292, "backend": "cuda", "own_frac": 1.0, "ref_ms": 3.73047, "cand_ms": 1.71054, "endpoint": "endpoint-14", "r_wall_s": 71.35, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 7069, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00042", "task": "L3_036", "split": "calib", "reader": "claude-opus-5-5", "arm": "reader_relative", "label": "reader_relative", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L3_036|claude-opus-5-5|1", "design": "f1", "success": true, "score": 2.1657, "status": "ok", "correct": true, "speedup": 2.1657, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 304.7, "calls": 3, "tokens_in": 28099, "tokens_out": 17412, "ctx_chars": 8756, "ctx_sha": "26dc0616aa00d951", "late_delivered": null, "t_end": 1791552702.981502, "backend": "cuda", "own_frac": 0.9998, "ref_ms": 37.24625, "cand_ms": 17.19817, "endpoint": "endpoint-04", "r_wall_s": 50.69, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 10439, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00080", "task": "L2_035", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "show_all", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_035|claude-opus-5-5|1", "design": "f1", "success": false, "score": 0.0, "status": "runtime_error", "correct": false, "speedup": 0.0, "infra": false, "infra_reruns": 0, "error": "OutOfMemoryError: CUDA out of memory. Tried to allocate 1024.00 MiB. GPU 0 has a total capacity of 39.49 GiB of which 677.12 MiB is free. Including non-PyTorch memory, this process has 2.23 GiB memory in use. Of the allocated memory 1.75 GiB is allocated by PyTorch, and 1.72 MiB is reserved by PyTor", "wall_s": 50.8, "calls": 2, "tokens_in": 13198, "tokens_out": 2545, "ctx_chars": 9835, "ctx_sha": "6aedb10d18ef2e36", "late_delivered": null, "t_end": 1791552718.907129, "backend": "triton", "own_frac": null, "ref_ms": null, "cand_ms": null, "endpoint": "endpoint-04", "r_wall_s": 7.69, "flags": null, "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 1588, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00057", "task": "L1_072", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "late_show_all", "delivery": "late", "rep": 0, "p_arm": 0.111111, "block": "L1_072|claude-opus-5-5|0", "design": "f1", "success": true, "score": 6.4062, "status": "ok", "correct": true, "speedup": 6.4062, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 213.1, "calls": 3, "tokens_in": 26150, "tokens_out": 14773, "ctx_chars": 11175, "ctx_sha": "91e4a03c424736b2", "late_delivered": true, "t_end": 1791552724.9235249, "backend": "cuda", "own_frac": 1.0, "ref_ms": 4.40699, "cand_ms": 0.68793, "endpoint": "endpoint-12", "r_wall_s": 36.07, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 8871, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00072", "task": "L2_025", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "show_all", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_025|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.3136, "status": "ok", "correct": true, "speedup": 1.3136, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 93.1, "calls": 2, "tokens_in": 13491, "tokens_out": 4269, "ctx_chars": 10087, "ctx_sha": "fc8a2930809fd512", "late_delivered": null, "t_end": 1791552726.1039505, "backend": "triton", "own_frac": 1.0, "ref_ms": 19.35409, "cand_ms": 14.73338, "endpoint": "endpoint-17", "r_wall_s": 30.68, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 3017, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00075", "task": "L2_008", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_numbers", "label": "no_numbers", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_008|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.0695, "status": "ok", "correct": true, "speedup": 1.0695, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 78.5, "calls": 2, "tokens_in": 14909, "tokens_out": 3342, "ctx_chars": 11057, "ctx_sha": "db6f9b6c0fb8c228", "late_delivered": null, "t_end": 1791552728.281493, "backend": "triton", "own_frac": 0.0237, "ref_ms": 12.70567, "cand_ms": 11.88043, "endpoint": "endpoint-09", "r_wall_s": 12.58, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 2259, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00071", "task": "L2_085", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "show_all", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_085|claude-opus-5-5|1", "design": "f1", "success": true, "score": 1.6083, "status": "ok", "correct": true, "speedup": 1.6083, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 103.5, "calls": 2, "tokens_in": 16970, "tokens_out": 5401, "ctx_chars": 11886, "ctx_sha": "8cbbd063f294068f", "late_delivered": null, "t_end": 1791552735.2216113, "backend": "triton", "own_frac": 0.2499, "ref_ms": 4.78912, "cand_ms": 2.97769, "endpoint": "endpoint-24", "r_wall_s": 45.19, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 3430, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00070", "task": "L3_046", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "late_show_all", "delivery": "late", "rep": 0, "p_arm": 0.111111, "block": "L3_046|claude-opus-5-5|0", "design": "f1", "success": true, "score": 2.3282, "status": "ok", "correct": true, "speedup": 2.3282, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 112.6, "calls": 2, "tokens_in": 12617, "tokens_out": 7550, "ctx_chars": 10384, "ctx_sha": "4573cb1c62d0cdf8", "late_delivered": true, "t_end": 1791552736.0701728, "backend": "triton", "own_frac": 0.4412, "ref_ms": 4.01368, "cand_ms": 1.72396, "endpoint": "endpoint-11", "r_wall_s": 32.59, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 5470, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00074", "task": "L1_065", "split": "calib", "reader": "claude-opus-5-5", "arm": "reader_relative", "label": "reader_relative", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L1_065|claude-opus-5-5|0", "design": "f1", "success": true, "score": 2.1938, "status": "ok", "correct": true, "speedup": 2.1938, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 98.0, "calls": 2, "tokens_in": 15547, "tokens_out": 2224, "ctx_chars": 11807, "ctx_sha": "04bab9b1ff7e3fde", "late_delivered": null, "t_end": 1791552739.4821076, "backend": "triton", "own_frac": 0.0001, "ref_ms": 44.3584, "cand_ms": 20.22002, "endpoint": "endpoint-07", "r_wall_s": 61.0, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 1195, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00067", "task": "L2_069", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "show_all", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_069|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.3246, "status": "ok", "correct": true, "speedup": 1.3246, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 130.9, "calls": 2, "tokens_in": 16281, "tokens_out": 4407, "ctx_chars": 12471, "ctx_sha": "4ddcc848d0210ca9", "late_delivered": null, "t_end": 1791552740.8382015, "backend": "triton", "own_frac": 0.9988, "ref_ms": 3.63628, "cand_ms": 2.74515, "endpoint": "endpoint-05", "r_wall_s": 60.12, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 3100, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00054", "task": "L3_049", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "late_show_all", "delivery": "late", "rep": 0, "p_arm": 0.111111, "block": "L3_049|claude-opus-5-5|0", "design": "f1", "success": true, "score": 8.8102, "status": "ok", "correct": true, "speedup": 8.8102, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 244.6, "calls": 4, "tokens_in": 31180, "tokens_out": 15415, "ctx_chars": 2425, "ctx_sha": "5c5e8811718de973", "late_delivered": true, "t_end": 1791552746.5268936, "backend": "cuda", "own_frac": 0.7727, "ref_ms": 12.01729, "cand_ms": 1.36403, "endpoint": "endpoint-16", "r_wall_s": 50.86, "flags": [], "submit_turn": 3, "turns": 4, "tech_src": "cuda", "tok_out_first": 4566, "stops": ["end_turn", "end_turn", "end_turn", "end_turn"], "first_check": "incorrect"} {"ep": "f1-00068", "task": "L1_050", "split": "calib", "reader": "claude-opus-5-5", "arm": "none", "label": "none", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L1_050|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.0823, "status": "ok", "correct": true, "speedup": 1.0823, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 134.5, "calls": 2, "tokens_in": 5468, "tokens_out": 5790, "ctx_chars": 0, "ctx_sha": "e3b0c44298fc1c14", "late_delivered": null, "t_end": 1791552747.2553287, "backend": "triton", "own_frac": 0.9988, "ref_ms": 5.34884, "cand_ms": 4.94212, "endpoint": "endpoint-24", "r_wall_s": 55.93, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 3841, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00062", "task": "L2_069", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_approach", "label": "no_approach", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_069|claude-opus-5-5|0", "design": "f1", "success": true, "score": 2.2781, "status": "ok", "correct": true, "speedup": 2.2781, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 181.5, "calls": 2, "tokens_in": 8681, "tokens_out": 8345, "ctx_chars": 3080, "ctx_sha": "7364d33fba159fed", "late_delivered": null, "t_end": 1791552749.1043477, "backend": "cuda", "own_frac": 1.0, "ref_ms": 3.89202, "cand_ms": 1.70847, "endpoint": "endpoint-01", "r_wall_s": 64.74, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 5932, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00065", "task": "L1_065", "split": "calib", "reader": "claude-opus-5-5", "arm": "frontier_gated", "label": "frontier_gated", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L1_065|claude-opus-5-5|0", "design": "f1", "success": true, "score": 3.2049, "status": "ok", "correct": true, "speedup": 3.2049, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 186.0, "calls": 2, "tokens_in": 9077, "tokens_out": 7007, "ctx_chars": 3256, "ctx_sha": "8fa366c59fab0c7f", "late_delivered": null, "t_end": 1791552767.1242166, "backend": "triton", "own_frac": 0.9997, "ref_ms": 44.2077, "cand_ms": 13.79377, "endpoint": "endpoint-18", "r_wall_s": 69.47, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 5094, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00073", "task": "L1_062", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "late_show_all", "delivery": "late", "rep": 1, "p_arm": 0.111111, "block": "L1_062|claude-opus-5-5|1", "design": "f1", "success": false, "score": 0.0, "status": "runtime_error", "correct": false, "speedup": 0.0, "infra": false, "infra_reruns": 0, "error": "OutOfMemoryError: CUDA out of memory. Tried to allocate 512.00 MiB. GPU 0 has a total capacity of 39.49 GiB of which 28.81 MiB is free. Including non-PyTorch memory, this process has 1.48 GiB memory in use. Of the allocated memory 1.00 GiB is allocated by PyTorch, and 1.65 MiB is reserved by PyTorch", "wall_s": 137.5, "calls": 2, "tokens_in": 11928, "tokens_out": 11082, "ctx_chars": 10864, "ctx_sha": "0b6222244e19e377", "late_delivered": true, "t_end": 1791552772.4637165, "backend": "cuda", "own_frac": null, "ref_ms": null, "cand_ms": null, "endpoint": "endpoint-13", "r_wall_s": 16.21, "flags": null, "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 7655, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00063", "task": "L3_038", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "show_all", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L3_038|claude-opus-5-5|1", "design": "f1", "success": true, "score": 3.7551, "status": "ok", "correct": true, "speedup": 3.7551, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 206.1, "calls": 3, "tokens_in": 27322, "tokens_out": 17880, "ctx_chars": 8224, "ctx_sha": "71fd1a97cbd02e3b", "late_delivered": null, "t_end": 1791552778.9164102, "backend": "cuda", "own_frac": 0.8443, "ref_ms": 79.07405, "cand_ms": 21.058, "endpoint": "endpoint-05", "r_wall_s": 19.36, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 11494, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00078", "task": "L2_035", "split": "calib", "reader": "claude-opus-5-5", "arm": "lessons_only", "label": "lessons_only", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_035|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.2692, "status": "ok", "correct": true, "speedup": 1.2692, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 123.2, "calls": 2, "tokens_in": 6938, "tokens_out": 4399, "ctx_chars": 1982, "ctx_sha": "54e16caf9bef9eec", "late_delivered": null, "t_end": 1791552781.4682112, "backend": "cuda", "own_frac": 0.0734, "ref_ms": 17.60852, "cand_ms": 13.87417, "endpoint": "endpoint-09", "r_wall_s": 42.25, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 2939, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00081", "task": "L2_052", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_numbers", "label": "no_numbers", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_052|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.3568, "status": "ok", "correct": true, "speedup": 1.3568, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 122.9, "calls": 2, "tokens_in": 15053, "tokens_out": 4426, "ctx_chars": 11127, "ctx_sha": "e2f40f0540528624", "late_delivered": null, "t_end": 1791552795.1952257, "backend": "triton", "own_frac": 0.1859, "ref_ms": 11.06103, "cand_ms": 8.15237, "endpoint": "endpoint-19", "r_wall_s": 59.4, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 2757, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00064", "task": "L1_072", "split": "calib", "reader": "claude-opus-5-5", "arm": "fg_target", "label": "fg_target", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L1_072|claude-opus-5-5|1", "design": "f1", "success": true, "score": 4.5607, "status": "ok", "correct": true, "speedup": 4.5607, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 223.0, "calls": 3, "tokens_in": 21656, "tokens_out": 18514, "ctx_chars": 2577, "ctx_sha": "6db86af015c9bcf3", "late_delivered": null, "t_end": 1791552796.0091646, "backend": "cuda", "own_frac": 1.0, "ref_ms": 4.39798, "cand_ms": 0.96433, "endpoint": "endpoint-01", "r_wall_s": 32.11, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 11029, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00069", "task": "L2_024", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "show_all", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_024|claude-opus-5-5|1", "design": "f1", "success": true, "score": 2.5373, "status": "ok", "correct": true, "speedup": 2.5373, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 180.5, "calls": 3, "tokens_in": 26703, "tokens_out": 12218, "ctx_chars": 10177, "ctx_sha": "cf1072ae0bde8a1b", "late_delivered": null, "t_end": 1791552801.1194246, "backend": "cuda", "own_frac": 1.0, "ref_ms": 2.38285, "cand_ms": 0.93913, "endpoint": "endpoint-24", "r_wall_s": 36.33, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 7022, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00066", "task": "L1_078", "split": "calib", "reader": "claude-opus-5-5", "arm": "frontier_gated", "label": "frontier_gated", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L1_078|claude-opus-5-5|0", "design": "f1", "success": true, "score": 3.0651, "status": "ok", "correct": true, "speedup": 3.0651, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 214.8, "calls": 2, "tokens_in": 9872, "tokens_out": 11140, "ctx_chars": 3256, "ctx_sha": "8fa366c59fab0c7f", "late_delivered": null, "t_end": 1791552822.443278, "backend": "cuda", "own_frac": 1.0, "ref_ms": 35.73608, "cand_ms": 11.65912, "endpoint": "endpoint-25", "r_wall_s": 71.31, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 8317, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00086", "task": "L3_046", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_approach", "label": "no_approach", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L3_046|claude-opus-5-5|0", "design": "f1", "success": false, "score": 0.0, "status": "runtime_error", "correct": false, "speedup": 0.0, "infra": false, "infra_reruns": 0, "error": "OutOfMemoryError: CUDA out of memory. Tried to allocate 800.00 MiB. GPU 0 has a total capacity of 39.49 GiB of which 284.81 MiB is free. Including non-PyTorch memory, this process has 1.05 GiB memory in use. Of the allocated memory 656.16 MiB is allocated by PyTorch, and 1.84 MiB is reserved by PyTo", "wall_s": 96.7, "calls": 2, "tokens_in": 10589, "tokens_out": 8533, "ctx_chars": 3288, "ctx_sha": "dc1a3e459ffe3f99", "late_delivered": null, "t_end": 1791552822.7665174, "backend": "triton", "own_frac": null, "ref_ms": null, "cand_ms": null, "endpoint": "endpoint-12", "r_wall_s": 5.09, "flags": null, "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 6554, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00083", "task": "L1_062", "split": "calib", "reader": "claude-opus-5-5", "arm": "lessons_only", "label": "lessons_only", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L1_062|claude-opus-5-5|0", "design": "f1", "success": false, "score": 0.4294, "status": "ok", "correct": true, "speedup": 0.4294, "infra": false, "infra_reruns": 0, "error": "ok", "wall_s": 150.4, "calls": 2, "tokens_in": 8116, "tokens_out": 7086, "ctx_chars": 1982, "ctx_sha": "54e16caf9bef9eec", "late_delivered": null, "t_end": 1791552853.342554, "backend": "triton", "own_frac": 0.9998, "ref_ms": 10.12066, "cand_ms": 23.57001, "endpoint": "endpoint-16", "r_wall_s": 62.45, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 5104, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00084", "task": "L2_052", "split": "calib", "reader": "claude-opus-5-5", "arm": "lessons_only", "label": "lessons_only", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_052|claude-opus-5-5|1", "design": "f1", "success": true, "score": 1.4064, "status": "ok", "correct": true, "speedup": 1.4064, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 147.0, "calls": 3, "tokens_in": 13262, "tokens_out": 6642, "ctx_chars": 1982, "ctx_sha": "54e16caf9bef9eec", "late_delivered": null, "t_end": 1791552865.95664, "backend": "triton", "own_frac": 0.1347, "ref_ms": 11.05744, "cand_ms": 7.86249, "endpoint": "endpoint-19", "r_wall_s": 57.18, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "triton", "tok_out_first": 3334, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "forbidden"} {"ep": "f1-00091", "task": "L2_052", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_approach", "label": "no_approach", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_052|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.403, "status": "ok", "correct": true, "speedup": 1.403, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 129.1, "calls": 2, "tokens_in": 7716, "tokens_out": 4632, "ctx_chars": 3080, "ctx_sha": "7364d33fba159fed", "late_delivered": null, "t_end": 1791552869.9283051, "backend": "triton", "own_frac": 0.1334, "ref_ms": 11.04014, "cand_ms": 7.8689, "endpoint": "endpoint-11", "r_wall_s": 53.39, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 3119, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00077", "task": "L2_067", "split": "calib", "reader": "claude-opus-5-5", "arm": "fg_target", "label": "fg_target", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_067|claude-opus-5-5|0", "design": "f1", "success": false, "score": 0.2366, "status": "ok", "correct": true, "speedup": 0.2366, "infra": false, "infra_reruns": 0, "error": "ok", "wall_s": 219.8, "calls": 4, "tokens_in": 29275, "tokens_out": 15370, "ctx_chars": 3639, "ctx_sha": "fc6e42d43dfb4e75", "late_delivered": null, "t_end": 1791552874.6307387, "backend": "cuda", "own_frac": 1.0, "ref_ms": 12.91233, "cand_ms": 54.56787, "endpoint": "endpoint-25", "r_wall_s": 43.51, "flags": [], "submit_turn": 3, "turns": 4, "tech_src": "cuda", "tok_out_first": 7613, "stops": ["end_turn", "end_turn", "end_turn", "end_turn"], "first_check": "forbidden"} {"ep": "f1-00092", "task": "L1_062", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "show_all", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L1_062|claude-opus-5-5|0", "design": "f1", "success": false, "score": 0.0, "status": "runtime_error", "correct": false, "speedup": 0.0, "infra": false, "infra_reruns": 0, "error": "OutOfMemoryError: CUDA out of memory. Tried to allocate 512.00 MiB. GPU 0 has a total capacity of 39.49 GiB of which 28.81 MiB is free. Including non-PyTorch memory, this process has 1.48 GiB memory in use. Of the allocated memory 1.00 GiB is allocated by PyTorch, and 1.65 MiB is reserved by PyTorch", "wall_s": 129.3, "calls": 2, "tokens_in": 15607, "tokens_out": 6125, "ctx_chars": 10864, "ctx_sha": "0b6222244e19e377", "late_delivered": null, "t_end": 1791552875.8630786, "backend": "triton", "own_frac": null, "ref_ms": null, "cand_ms": null, "endpoint": "endpoint-13", "r_wall_s": 7.08, "flags": null, "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 4284, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00097", "task": "L2_057", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "show_all", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_057|claude-opus-5-5|0", "design": "f1", "success": true, "score": 2.1672, "status": "ok", "correct": true, "speedup": 2.1672, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 110.2, "calls": 3, "tokens_in": 20386, "tokens_out": 2547, "ctx_chars": 9855, "ctx_sha": "1a6a80cca09cee48", "late_delivered": null, "t_end": 1791552889.0862741, "backend": "triton", "own_frac": 0.2378, "ref_ms": 6.30132, "cand_ms": 2.90756, "endpoint": "endpoint-09", "r_wall_s": 54.37, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "triton", "tok_out_first": 1205, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "runtime_error"} {"ep": "f1-00082", "task": "L3_042", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "late_show_all", "delivery": "late", "rep": 0, "p_arm": 0.111111, "block": "L3_042|claude-opus-5-5|0", "design": "f1", "success": true, "score": 3.9263, "status": "ok", "correct": true, "speedup": 3.9263, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 197.5, "calls": 3, "tokens_in": 24275, "tokens_out": 16170, "ctx_chars": 8224, "ctx_sha": "71fd1a97cbd02e3b", "late_delivered": true, "t_end": 1791552897.9790506, "backend": "cuda", "own_frac": 0.8371, "ref_ms": 60.53061, "cand_ms": 15.4169, "endpoint": "endpoint-23", "r_wall_s": 21.5, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 9166, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00090", "task": "L2_074", "split": "calib", "reader": "claude-opus-5-5", "arm": "none", "label": "none", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_074|claude-opus-5-5|0", "design": "f1", "success": true, "score": 9.9578, "status": "ok", "correct": true, "speedup": 9.9578, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 161.3, "calls": 3, "tokens_in": 14303, "tokens_out": 11489, "ctx_chars": 0, "ctx_sha": "e3b0c44298fc1c14", "late_delivered": null, "t_end": 1791552900.7410662, "backend": "cuda", "own_frac": 1.0, "ref_ms": 11.02778, "cand_ms": 1.10746, "endpoint": "endpoint-04", "r_wall_s": 26.59, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 6360, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00108", "task": "L2_079", "split": "calib", "reader": "claude-opus-5-5", "arm": "frontier_gated", "label": "frontier_gated", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_079|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.4642, "status": "ok", "correct": true, "speedup": 1.4642, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 57.7, "calls": 2, "tokens_in": 8323, "tokens_out": 4027, "ctx_chars": 3639, "ctx_sha": "fc6e42d43dfb4e75", "late_delivered": null, "t_end": 1791552933.5969057, "backend": "triton", "own_frac": 0.1055, "ref_ms": 2.17334, "cand_ms": 1.48434, "endpoint": "endpoint-14", "r_wall_s": 11.98, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 2551, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00085", "task": "L2_082", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "late_show_all", "delivery": "late", "rep": 1, "p_arm": 0.111111, "block": "L2_082|claude-opus-5-5|1", "design": "f1", "success": true, "score": 3.4388, "status": "ok", "correct": true, "speedup": 3.4388, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 208.9, "calls": 3, "tokens_in": 23932, "tokens_out": 11788, "ctx_chars": 10743, "ctx_sha": "30d3b01036fcb086", "late_delivered": true, "t_end": 1791552933.871103, "backend": "cuda", "own_frac": 1.0, "ref_ms": 19.55898, "cand_ms": 5.68777, "endpoint": "endpoint-16", "r_wall_s": 63.5, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 6187, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00079", "task": "L1_085", "split": "calib", "reader": "claude-opus-5-5", "arm": "none", "label": "none", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L1_085|claude-opus-5-5|0", "design": "f1", "success": true, "score": 3.0164, "status": "ok", "correct": true, "speedup": 3.0164, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 271.6, "calls": 2, "tokens_in": 8161, "tokens_out": 8706, "ctx_chars": 0, "ctx_sha": "e3b0c44298fc1c14", "late_delivered": null, "t_end": 1791552936.6964161, "backend": "cuda", "own_frac": 1.0, "ref_ms": 3.73577, "cand_ms": 1.23849, "endpoint": "endpoint-01", "r_wall_s": 60.72, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 5626, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00087", "task": "L3_004", "split": "calib", "reader": "claude-opus-5-5", "arm": "fg_target", "label": "fg_target", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L3_004|claude-opus-5-5|0", "design": "f1", "success": true, "score": 3.7295, "status": "ok", "correct": true, "speedup": 3.7295, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 215.0, "calls": 3, "tokens_in": 22871, "tokens_out": 15382, "ctx_chars": 2759, "ctx_sha": "e8b611ac667b08e3", "late_delivered": null, "t_end": 1791552943.3045707, "backend": "cuda", "own_frac": 1.0, "ref_ms": 2.30286, "cand_ms": 0.61746, "endpoint": "endpoint-10", "r_wall_s": 19.75, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 6887, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00102", "task": "L1_078", "split": "calib", "reader": "claude-opus-5-5", "arm": "lessons_only", "label": "lessons_only", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L1_078|claude-opus-5-5|0", "design": "f1", "success": false, "score": 0.0, "status": "runtime_error", "correct": false, "speedup": 0.0, "infra": false, "infra_reruns": 0, "error": "OutOfMemoryError: CUDA out of memory. Tried to allocate 1024.00 MiB. GPU 0 has a total capacity of 39.49 GiB of which 108.81 MiB is free. Including non-PyTorch memory, this process has 1.16 GiB memory in use. Of the allocated memory 768.08 MiB is allocated by PyTorch, and 1.92 MiB is reserved by PyT", "wall_s": 121.5, "calls": 2, "tokens_in": 7942, "tokens_out": 5874, "ctx_chars": 1982, "ctx_sha": "54e16caf9bef9eec", "late_delivered": null, "t_end": 1791552943.9556181, "backend": "triton", "own_frac": null, "ref_ms": null, "cand_ms": null, "endpoint": "endpoint-13", "r_wall_s": 6.8, "flags": null, "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 3883, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00089", "task": "L2_069", "split": "calib", "reader": "claude-opus-5-5", "arm": "none", "label": "none", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_069|claude-opus-5-5|1", "design": "f1", "success": true, "score": 2.6683, "status": "ok", "correct": true, "speedup": 2.6683, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 209.3, "calls": 3, "tokens_in": 13087, "tokens_out": 11129, "ctx_chars": 0, "ctx_sha": "e3b0c44298fc1c14", "late_delivered": null, "t_end": 1791552945.3382008, "backend": "cuda", "own_frac": 1.0, "ref_ms": 3.64383, "cand_ms": 1.3656, "endpoint": "endpoint-11", "r_wall_s": 58.85, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 6377, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00094", "task": "L3_042", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "show_all", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L3_042|claude-opus-5-5|0", "design": "f1", "success": true, "score": 3.2635, "status": "ok", "correct": true, "speedup": 3.2635, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 201.8, "calls": 3, "tokens_in": 28243, "tokens_out": 16708, "ctx_chars": 8224, "ctx_sha": "71fd1a97cbd02e3b", "late_delivered": null, "t_end": 1791552950.9068215, "backend": "cuda", "own_frac": 0.8532, "ref_ms": 60.64549, "cand_ms": 18.58316, "endpoint": "endpoint-25", "r_wall_s": 19.51, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 9489, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00100", "task": "L1_072", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "late_show_all", "delivery": "late", "rep": 1, "p_arm": 0.111111, "block": "L1_072|claude-opus-5-5|1", "design": "f1", "success": true, "score": 8.5653, "status": "ok", "correct": true, "speedup": 8.5653, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 165.2, "calls": 2, "tokens_in": 12506, "tokens_out": 10827, "ctx_chars": 11175, "ctx_sha": "91e4a03c424736b2", "late_delivered": true, "t_end": 1791552961.2247477, "backend": "cuda", "own_frac": 1.0, "ref_ms": 4.40637, "cand_ms": 0.51444, "endpoint": "endpoint-17", "r_wall_s": 29.57, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 7944, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00103", "task": "L1_071", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "late_show_all", "delivery": "late", "rep": 0, "p_arm": 0.111111, "block": "L1_071|claude-opus-5-5|0", "design": "f1", "success": false, "score": 0.0, "status": "runtime_error", "correct": false, "speedup": 0.0, "infra": false, "infra_reruns": 0, "error": "OutOfMemoryError: CUDA out of memory. Tried to allocate 1024.00 MiB. GPU 0 has a total capacity of 39.49 GiB of which 104.81 MiB is free. Including non-PyTorch memory, this process has 1.16 GiB memory in use. Of the allocated memory 768.04 MiB is allocated by PyTorch, and 1.96 MiB is reserved by PyT", "wall_s": 142.1, "calls": 2, "tokens_in": 12352, "tokens_out": 11526, "ctx_chars": 11175, "ctx_sha": "91e4a03c424736b2", "late_delivered": true, "t_end": 1791552964.8556097, "backend": "cuda", "own_frac": null, "ref_ms": null, "cand_ms": null, "endpoint": "endpoint-13", "r_wall_s": 16.06, "flags": null, "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 7910, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00104", "task": "L2_057", "split": "calib", "reader": "claude-opus-5-5", "arm": "reader_relative", "label": "reader_relative", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_057|claude-opus-5-5|1", "design": "f1", "success": true, "score": 2.1695, "status": "ok", "correct": true, "speedup": 2.1695, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 111.6, "calls": 2, "tokens_in": 13172, "tokens_out": 1808, "ctx_chars": 10481, "ctx_sha": "7d65586d1263a6d9", "late_delivered": null, "t_end": 1791552964.915918, "backend": "triton", "own_frac": 0.2423, "ref_ms": 6.29897, "cand_ms": 2.90344, "endpoint": "endpoint-26", "r_wall_s": 60.0, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 1244, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00095", "task": "L2_074", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "late_show_all", "delivery": "late", "rep": 0, "p_arm": 0.111111, "block": "L2_074|claude-opus-5-5|0", "design": "f1", "success": true, "score": 8.1183, "status": "ok", "correct": true, "speedup": 8.1183, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 200.0, "calls": 3, "tokens_in": 25280, "tokens_out": 13968, "ctx_chars": 10364, "ctx_sha": "19f3ac1a7ff915f6", "late_delivered": true, "t_end": 1791552967.0916054, "backend": "cuda", "own_frac": 1.0, "ref_ms": 5.90949, "cand_ms": 0.72792, "endpoint": "endpoint-14", "r_wall_s": 22.31, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 7598, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00093", "task": "L2_018", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "late_show_all", "delivery": "late", "rep": 1, "p_arm": 0.111111, "block": "L2_018|claude-opus-5-5|1", "design": "f1", "success": true, "score": 3.1365, "status": "ok", "correct": true, "speedup": 3.1365, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 230.4, "calls": 4, "tokens_in": 35566, "tokens_out": 15397, "ctx_chars": 9898, "ctx_sha": "eb48d1e5953cd4b8", "late_delivered": true, "t_end": 1791552977.6750064, "backend": "cuda", "own_frac": 0.1016, "ref_ms": 7.99151, "cand_ms": 2.54793, "endpoint": "endpoint-19", "r_wall_s": 27.83, "flags": [], "submit_turn": 3, "turns": 4, "tech_src": "cuda", "tok_out_first": 3629, "stops": ["end_turn", "end_turn", "end_turn", "end_turn"], "first_check": "incorrect"} {"ep": "f1-00088", "task": "L1_050", "split": "calib", "reader": "claude-opus-5-5", "arm": "fg_target", "label": "fg_target", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L1_050|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.1146, "status": "ok", "correct": true, "speedup": 1.1146, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 265.4, "calls": 3, "tokens_in": 19959, "tokens_out": 13257, "ctx_chars": 6597, "ctx_sha": "661091d9e159f800", "late_delivered": null, "t_end": 1791553000.6275268, "backend": "cuda", "own_frac": 0.9992, "ref_ms": 5.2863, "cand_ms": 4.74262, "endpoint": "endpoint-18", "r_wall_s": 81.25, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 8675, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00101", "task": "L2_082", "split": "calib", "reader": "claude-opus-5-5", "arm": "frontier_gated", "label": "frontier_gated", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_082|claude-opus-5-5|0", "design": "f1", "success": true, "score": 4.4135, "status": "ok", "correct": true, "speedup": 4.4135, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 201.7, "calls": 3, "tokens_in": 18589, "tokens_out": 14052, "ctx_chars": 3639, "ctx_sha": "fc6e42d43dfb4e75", "late_delivered": null, "t_end": 1791553002.791538, "backend": "cuda", "own_frac": 0.9991, "ref_ms": 19.51465, "cand_ms": 4.42158, "endpoint": "endpoint-17", "r_wall_s": 38.17, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 8301, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00115", "task": "L2_085", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "show_all", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_085|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.6077, "status": "ok", "correct": true, "speedup": 1.6077, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 62.7, "calls": 2, "tokens_in": 16740, "tokens_out": 4702, "ctx_chars": 11886, "ctx_sha": "8cbbd063f294068f", "late_delivered": null, "t_end": 1791553006.0495725, "backend": "triton", "own_frac": 0.2466, "ref_ms": 4.72456, "cand_ms": 2.93865, "endpoint": "endpoint-05", "r_wall_s": 14.73, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 2892, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00098", "task": "L3_036", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "late_show_all", "delivery": "late", "rep": 1, "p_arm": 0.111111, "block": "L3_036|claude-opus-5-5|1", "design": "f1", "success": true, "score": 5.9658, "status": "ok", "correct": true, "speedup": 5.9658, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 225.7, "calls": 3, "tokens_in": 24922, "tokens_out": 19362, "ctx_chars": 8224, "ctx_sha": "71fd1a97cbd02e3b", "late_delivered": true, "t_end": 1791553007.1407797, "backend": "cuda", "own_frac": 0.9997, "ref_ms": 32.23413, "cand_ms": 5.40316, "endpoint": "endpoint-04", "r_wall_s": 20.09, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 11911, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00118", "task": "L2_085", "split": "calib", "reader": "claude-opus-5-5", "arm": "frontier_gated", "label": "frontier_gated", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_085|claude-opus-5-5|1", "design": "f1", "success": true, "score": 1.3887, "status": "ok", "correct": true, "speedup": 1.3887, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 77.0, "calls": 2, "tokens_in": 12854, "tokens_out": 5624, "ctx_chars": 7376, "ctx_sha": "cc69cebe7b058084", "late_delivered": null, "t_end": 1791553027.9455168, "backend": "triton", "own_frac": 0.3701, "ref_ms": 4.73422, "cand_ms": 3.40909, "endpoint": "endpoint-04", "r_wall_s": 16.13, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 3628, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00099", "task": "L3_039", "split": "calib", "reader": "claude-opus-5-5", "arm": "reader_relative", "label": "reader_relative", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L3_039|claude-opus-5-5|1", "design": "f1", "success": true, "score": 6.0441, "status": "ok", "correct": true, "speedup": 6.0441, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 243.0, "calls": 3, "tokens_in": 30509, "tokens_out": 22025, "ctx_chars": 8756, "ctx_sha": "26dc0616aa00d951", "late_delivered": null, "t_end": 1791553038.1859086, "backend": "cuda", "own_frac": 0.9809, "ref_ms": 31.41798, "cand_ms": 5.19808, "endpoint": "endpoint-13", "r_wall_s": 20.03, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 13557, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00120", "task": "L2_079", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "show_all", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_079|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.4757, "status": "ok", "correct": true, "speedup": 1.4757, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 75.6, "calls": 2, "tokens_in": 14755, "tokens_out": 4054, "ctx_chars": 10508, "ctx_sha": "4fa7857ea4c0dcf4", "late_delivered": null, "t_end": 1791553040.4794304, "backend": "triton", "own_frac": 0.1021, "ref_ms": 2.18535, "cand_ms": 1.48085, "endpoint": "endpoint-08", "r_wall_s": 20.66, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 2551, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00123", "task": "L2_008", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_approach", "label": "no_approach", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_008|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.0518, "status": "ok", "correct": true, "speedup": 1.0518, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 72.8, "calls": 2, "tokens_in": 7902, "tokens_out": 4858, "ctx_chars": 3080, "ctx_sha": "7364d33fba159fed", "late_delivered": null, "t_end": 1791553050.5215745, "backend": "triton", "own_frac": 0.0467, "ref_ms": 12.68918, "cand_ms": 12.06451, "endpoint": "endpoint-01", "r_wall_s": 12.24, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 3686, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00113", "task": "L2_035", "split": "calib", "reader": "claude-opus-5-5", "arm": "none", "label": "none", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_035|claude-opus-5-5|1", "design": "f1", "success": true, "score": 1.2643, "status": "ok", "correct": true, "speedup": 1.2643, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 119.2, "calls": 2, "tokens_in": 5302, "tokens_out": 3859, "ctx_chars": 0, "ctx_sha": "e3b0c44298fc1c14", "late_delivered": null, "t_end": 1791553053.0870457, "backend": "cuda", "own_frac": 0.0738, "ref_ms": 17.41017, "cand_ms": 13.77026, "endpoint": "endpoint-05", "r_wall_s": 41.67, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 2447, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00116", "task": "L1_071", "split": "calib", "reader": "claude-opus-5-5", "arm": "fg_target", "label": "fg_target", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L1_071|claude-opus-5-5|1", "design": "f1", "success": false, "score": 0.0, "status": "runtime_error", "correct": false, "speedup": 0.0, "infra": false, "infra_reruns": 0, "error": "AcceleratorError: CUDA error: out of memory\nSearch for `cudaErrorMemoryAllocation' in https://docs.nvidia.com/cuda/cuda-runtime-api/group__CUDART__TYPES.html for more information.\nCUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.\n", "wall_s": 109.7, "calls": 2, "tokens_in": 8320, "tokens_out": 6579, "ctx_chars": 2577, "ctx_sha": "6db86af015c9bcf3", "late_delivered": null, "t_end": 1791553053.606119, "backend": "triton", "own_frac": null, "ref_ms": null, "cand_ms": null, "endpoint": "endpoint-13", "r_wall_s": 7.08, "flags": null, "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 4659, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00107", "task": "L2_024", "split": "calib", "reader": "claude-opus-5-5", "arm": "fg_target", "label": "fg_target", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_024|claude-opus-5-5|1", "design": "f1", "success": true, "score": 2.3048, "status": "ok", "correct": true, "speedup": 2.3048, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 185.3, "calls": 2, "tokens_in": 8868, "tokens_out": 7606, "ctx_chars": 3639, "ctx_sha": "fc6e42d43dfb4e75", "late_delivered": null, "t_end": 1791553059.939321, "backend": "cuda", "own_frac": 0.9951, "ref_ms": 2.37372, "cand_ms": 1.02989, "endpoint": "endpoint-25", "r_wall_s": 68.13, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 5456, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00114", "task": "L2_025", "split": "calib", "reader": "claude-opus-5-5", "arm": "lessons_only", "label": "lessons_only", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_025|claude-opus-5-5|1", "design": "f1", "success": true, "score": 2.0697, "status": "ok", "correct": true, "speedup": 2.0697, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 139.9, "calls": 2, "tokens_in": 7842, "tokens_out": 11264, "ctx_chars": 1982, "ctx_sha": "54e16caf9bef9eec", "late_delivered": null, "t_end": 1791553076.6341736, "backend": "cuda", "own_frac": 1.0, "ref_ms": 19.24482, "cand_ms": 9.29831, "endpoint": "endpoint-10", "r_wall_s": 21.46, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 8699, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00110", "task": "L2_082", "split": "calib", "reader": "claude-opus-5-5", "arm": "fg_target", "label": "fg_target", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_082|claude-opus-5-5|0", "design": "f1", "success": true, "score": 4.2064, "status": "ok", "correct": true, "speedup": 4.2064, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 183.8, "calls": 3, "tokens_in": 17864, "tokens_out": 11616, "ctx_chars": 3639, "ctx_sha": "fc6e42d43dfb4e75", "late_delivered": null, "t_end": 1791553081.771669, "backend": "cuda", "own_frac": 1.0, "ref_ms": 19.56584, "cand_ms": 4.65142, "endpoint": "endpoint-16", "r_wall_s": 32.19, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 6544, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00111", "task": "L1_077", "split": "calib", "reader": "claude-opus-5-5", "arm": "reader_relative", "label": "reader_relative", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L1_077|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.9832, "status": "ok", "correct": true, "speedup": 1.9832, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 190.0, "calls": 2, "tokens_in": 16851, "tokens_out": 7142, "ctx_chars": 11807, "ctx_sha": "a79e3b8677344448", "late_delivered": null, "t_end": 1791553090.7750661, "backend": "triton", "own_frac": 0.2213, "ref_ms": 8.61361, "cand_ms": 4.34322, "endpoint": "endpoint-09", "r_wall_s": 62.87, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 5003, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00124", "task": "L2_085", "split": "calib", "reader": "claude-opus-5-5", "arm": "fg_target", "label": "fg_target", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_085|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.6117, "status": "ok", "correct": true, "speedup": 1.6117, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 91.8, "calls": 2, "tokens_in": 12695, "tokens_out": 5144, "ctx_chars": 7376, "ctx_sha": "cc69cebe7b058084", "late_delivered": null, "t_end": 1791553092.4597495, "backend": "triton", "own_frac": 0.252, "ref_ms": 4.74374, "cand_ms": 2.94328, "endpoint": "endpoint-01", "r_wall_s": 29.6, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 3261, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00106", "task": "L2_073", "split": "calib", "reader": "claude-opus-5-5", "arm": "none", "label": "none", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_073|claude-opus-5-5|1", "design": "f1", "success": true, "score": 2.1464, "status": "ok", "correct": true, "speedup": 2.1464, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 226.9, "calls": 2, "tokens_in": 7971, "tokens_out": 15258, "ctx_chars": 0, "ctx_sha": "e3b0c44298fc1c14", "late_delivered": null, "t_end": 1791553096.8692307, "backend": "cuda", "own_frac": 0.9989, "ref_ms": 5.61184, "cand_ms": 2.61452, "endpoint": "endpoint-18", "r_wall_s": 76.08, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 11005, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00121", "task": "L2_025", "split": "calib", "reader": "claude-opus-5-5", "arm": "fg_target", "label": "fg_target", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_025|claude-opus-5-5|0", "design": "f1", "success": true, "score": 2.7305, "status": "ok", "correct": true, "speedup": 2.7305, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 143.3, "calls": 2, "tokens_in": 7953, "tokens_out": 9170, "ctx_chars": 3639, "ctx_sha": "fc6e42d43dfb4e75", "late_delivered": null, "t_end": 1791553108.1794531, "backend": "triton", "own_frac": 0.9995, "ref_ms": 19.32879, "cand_ms": 7.07892, "endpoint": "endpoint-07", "r_wall_s": 27.69, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 7772, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00134", "task": "L2_067", "split": "calib", "reader": "claude-opus-5-5", "arm": "reader_relative", "label": "reader_relative", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_067|claude-opus-5-5|1", "design": "f1", "success": true, "score": 1.2708, "status": "ok", "correct": true, "speedup": 1.2708, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 53.4, "calls": 2, "tokens_in": 14465, "tokens_out": 2093, "ctx_chars": 11649, "ctx_sha": "4bc9a6c4c4f94c6f", "late_delivered": null, "t_end": 1791553113.30307, "backend": "triton", "own_frac": 0.1628, "ref_ms": 12.85947, "cand_ms": 10.11923, "endpoint": "endpoint-25", "r_wall_s": 12.31, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 1473, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00096", "task": "L3_036", "split": "calib", "reader": "claude-opus-5-5", "arm": "none", "label": "none", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L3_036|claude-opus-5-5|1", "design": "f1", "success": true, "score": 4.7955, "status": "ok", "correct": true, "speedup": 4.7955, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 343.1, "calls": 5, "tokens_in": 42677, "tokens_out": 31409, "ctx_chars": 0, "ctx_sha": "e3b0c44298fc1c14", "late_delivered": null, "t_end": 1791553115.5701683, "backend": "cuda", "own_frac": 0.9997, "ref_ms": 32.62424, "cand_ms": 6.80304, "endpoint": "endpoint-24", "r_wall_s": 29.93, "flags": [], "submit_turn": 4, "turns": 5, "tech_src": "cuda", "tok_out_first": 9807, "stops": [null, "end_turn", "end_turn", "end_turn", "end_turn"], "first_check": "no_code"} {"ep": "f1-00125", "task": "L3_004", "split": "calib", "reader": "claude-opus-5-5", "arm": "lessons_only", "label": "lessons_only", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L3_004|claude-opus-5-5|1", "design": "f1", "success": true, "score": 4.481, "status": "ok", "correct": true, "speedup": 4.481, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 124.4, "calls": 2, "tokens_in": 8757, "tokens_out": 7539, "ctx_chars": 1982, "ctx_sha": "54e16caf9bef9eec", "late_delivered": null, "t_end": 1791553127.1845129, "backend": "cuda", "own_frac": 0.7858, "ref_ms": 2.29596, "cand_ms": 0.51238, "endpoint": "endpoint-12", "r_wall_s": 20.45, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 4937, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00109", "task": "L1_080", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "late_show_all", "delivery": "late", "rep": 0, "p_arm": 0.111111, "block": "L1_080|claude-opus-5-5|0", "design": "f1", "success": false, "score": 0.8016, "status": "ok", "correct": true, "speedup": 0.8016, "infra": false, "infra_reruns": 0, "error": "ok", "wall_s": 239.4, "calls": 3, "tokens_in": 26980, "tokens_out": 17579, "ctx_chars": 10864, "ctx_sha": "0b6222244e19e377", "late_delivered": true, "t_end": 1791553128.4366407, "backend": "cuda", "own_frac": 1.0, "ref_ms": 20.18585, "cand_ms": 25.1826, "endpoint": "endpoint-08", "r_wall_s": 61.09, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 10275, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "runtime_error"} {"ep": "f1-00129", "task": "L2_011", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_numbers", "label": "no_numbers", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_011|claude-opus-5-5|1", "design": "f1", "success": true, "score": 4.1162, "status": "ok", "correct": true, "speedup": 4.1162, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 91.6, "calls": 2, "tokens_in": 17495, "tokens_out": 6252, "ctx_chars": 12347, "ctx_sha": "89a205c70b5e9cde", "late_delivered": null, "t_end": 1791553129.8232312, "backend": "triton", "own_frac": 0.1528, "ref_ms": 27.80514, "cand_ms": 6.75499, "endpoint": "endpoint-18", "r_wall_s": 21.61, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 4077, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00105", "task": "L3_039", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "late_show_all", "delivery": "late", "rep": 0, "p_arm": 0.111111, "block": "L3_039|claude-opus-5-5|0", "design": "f1", "success": true, "score": 6.6951, "status": "ok", "correct": true, "speedup": 6.6951, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 268.1, "calls": 3, "tokens_in": 25933, "tokens_out": 22486, "ctx_chars": 8224, "ctx_sha": "71fd1a97cbd02e3b", "late_delivered": true, "t_end": 1791553134.053024, "backend": "cuda", "own_frac": 0.9768, "ref_ms": 31.57042, "cand_ms": 4.71543, "endpoint": "endpoint-07", "r_wall_s": 21.3, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 14281, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00112", "task": "L1_072", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_numbers", "label": "no_numbers", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L1_072|claude-opus-5-5|0", "design": "f1", "success": true, "score": 6.5421, "status": "ok", "correct": true, "speedup": 6.5421, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 209.0, "calls": 3, "tokens_in": 31251, "tokens_out": 15326, "ctx_chars": 11170, "ctx_sha": "ee0e99e621c34bdc", "late_delivered": null, "t_end": 1791553142.5678208, "backend": "cuda", "own_frac": 1.0, "ref_ms": 4.37781, "cand_ms": 0.66918, "endpoint": "endpoint-26", "r_wall_s": 28.6, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 8527, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00117", "task": "L2_025", "split": "calib", "reader": "claude-opus-5-5", "arm": "frontier_gated", "label": "frontier_gated", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_025|claude-opus-5-5|1", "design": "f1", "success": true, "score": 2.09, "status": "ok", "correct": true, "speedup": 2.09, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 204.5, "calls": 3, "tokens_in": 20655, "tokens_out": 12084, "ctx_chars": 5671, "ctx_sha": "7f0e3598c3810513", "late_delivered": null, "t_end": 1791553149.8404229, "backend": "cuda", "own_frac": 1.0, "ref_ms": 19.18232, "cand_ms": 9.17831, "endpoint": "endpoint-19", "r_wall_s": 22.99, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 7009, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00130", "task": "L1_085", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "show_all", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L1_085|claude-opus-5-5|0", "design": "f1", "success": true, "score": 2.2128, "status": "ok", "correct": true, "speedup": 2.2128, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 125.4, "calls": 2, "tokens_in": 15830, "tokens_out": 3420, "ctx_chars": 10864, "ctx_sha": "46fefeb0053babb6", "late_delivered": null, "t_end": 1791553165.9046202, "backend": "triton", "own_frac": 1.0, "ref_ms": 4.15242, "cand_ms": 1.87658, "endpoint": "endpoint-05", "r_wall_s": 68.48, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 1823, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00126", "task": "L1_080", "split": "calib", "reader": "claude-opus-5-5", "arm": "reader_relative", "label": "reader_relative", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L1_080|claude-opus-5-5|1", "design": "f1", "success": true, "score": 1.2571, "status": "ok", "correct": true, "speedup": 1.2571, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 165.0, "calls": 2, "tokens_in": 16034, "tokens_out": 6724, "ctx_chars": 11490, "ctx_sha": "643e4c44bbeb656a", "late_delivered": null, "t_end": 1791553171.0089614, "backend": "triton", "own_frac": 0.9997, "ref_ms": 20.21959, "cand_ms": 16.08381, "endpoint": "endpoint-04", "r_wall_s": 70.38, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 4889, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00132", "task": "L1_085", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_numbers", "label": "no_numbers", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L1_085|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.3211, "status": "ok", "correct": true, "speedup": 1.3211, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 131.4, "calls": 2, "tokens_in": 15871, "tokens_out": 3866, "ctx_chars": 10859, "ctx_sha": "cd499fa09a3829ac", "late_delivered": null, "t_end": 1791553184.4772148, "backend": "triton", "own_frac": 1.0, "ref_ms": 3.73281, "cand_ms": 2.82555, "endpoint": "endpoint-11", "r_wall_s": 75.28, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 2303, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00138", "task": "L2_011", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_numbers", "label": "no_numbers", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_011|claude-opus-5-5|0", "design": "f1", "success": true, "score": 4.4376, "status": "ok", "correct": true, "speedup": 4.4376, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 97.6, "calls": 2, "tokens_in": 17963, "tokens_out": 7564, "ctx_chars": 12347, "ctx_sha": "89a205c70b5e9cde", "late_delivered": null, "t_end": 1791553190.0208223, "backend": "triton", "own_frac": 0.0903, "ref_ms": 27.8297, "cand_ms": 6.27132, "endpoint": "endpoint-18", "r_wall_s": 20.98, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 4995, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00131", "task": "L3_039", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "show_all", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L3_039|claude-opus-5-5|1", "design": "f1", "success": true, "score": 2.2325, "status": "ok", "correct": true, "speedup": 2.2325, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 140.0, "calls": 2, "tokens_in": 14186, "tokens_out": 11222, "ctx_chars": 8224, "ctx_sha": "71fd1a97cbd02e3b", "late_delivered": null, "t_end": 1791553190.5135, "backend": "cuda", "own_frac": 0.9465, "ref_ms": 31.61624, "cand_ms": 14.16153, "endpoint": "endpoint-13", "r_wall_s": 20.58, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 8689, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00127", "task": "L3_042", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_numbers", "label": "no_numbers", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L3_042|claude-opus-5-5|1", "design": "f1", "success": true, "score": 3.574, "status": "ok", "correct": true, "speedup": 3.574, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 183.8, "calls": 3, "tokens_in": 27791, "tokens_out": 15735, "ctx_chars": 8204, "ctx_sha": "83777d4de17c3440", "late_delivered": null, "t_end": 1791553190.9365697, "backend": "cuda", "own_frac": 0.8389, "ref_ms": 60.88377, "cand_ms": 17.03528, "endpoint": "endpoint-07", "r_wall_s": 20.14, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 8774, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00119", "task": "L3_008", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "late_show_all", "delivery": "late", "rep": 0, "p_arm": 0.111111, "block": "L3_008|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.5275, "status": "ok", "correct": true, "speedup": 1.5275, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 234.4, "calls": 2, "tokens_in": 14497, "tokens_out": 21977, "ctx_chars": 6513, "ctx_sha": "1aa698d3a9589f25", "late_delivered": true, "t_end": 1791553195.6377223, "backend": "cuda", "own_frac": 0.1419, "ref_ms": 5.26398, "cand_ms": 3.4461, "endpoint": "endpoint-19", "r_wall_s": 31.89, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 14495, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00076", "task": "L3_024", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_numbers", "label": "no_numbers", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L3_024|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.3979, "status": "ok", "correct": true, "speedup": 1.3979, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 552.6, "calls": 5, "tokens_in": 90787, "tokens_out": 44590, "ctx_chars": 6490, "ctx_sha": "00c26ca778818693", "late_delivered": null, "t_end": 1791553204.4363508, "backend": "cuda", "own_frac": 0.0352, "ref_ms": 3.52859, "cand_ms": 2.5242, "endpoint": "endpoint-23", "r_wall_s": 20.48, "flags": [], "submit_turn": 4, "turns": 5, "tech_src": "cuda", "tok_out_first": 17635, "stops": ["end_turn", "end_turn", "end_turn", "end_turn", "end_turn"], "first_check": "runtime_error"} {"ep": "f1-00136", "task": "L2_079", "split": "calib", "reader": "claude-opus-5-5", "arm": "frontier_gated", "label": "frontier_gated", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_079|claude-opus-5-5|1", "design": "f1", "success": false, "score": 0.1333, "status": "ok", "correct": true, "speedup": 0.1333, "infra": false, "infra_reruns": 0, "error": "ok", "wall_s": 127.1, "calls": 2, "tokens_in": 9009, "tokens_out": 6514, "ctx_chars": 3639, "ctx_sha": "fc6e42d43dfb4e75", "late_delivered": null, "t_end": 1791553208.8422453, "backend": "triton", "own_frac": 0.9999, "ref_ms": 2.18115, "cand_ms": 16.36459, "endpoint": "endpoint-04", "r_wall_s": 30.82, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 4452, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00128", "task": "L3_046", "split": "calib", "reader": "claude-opus-5-5", "arm": "none", "label": "none", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L3_046|claude-opus-5-5|1", "design": "f1", "success": true, "score": 2.2643, "status": "ok", "correct": true, "speedup": 2.2643, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 191.6, "calls": 2, "tokens_in": 9583, "tokens_out": 12721, "ctx_chars": 0, "ctx_sha": "e3b0c44298fc1c14", "late_delivered": null, "t_end": 1791553219.520478, "backend": "cuda", "own_frac": 0.4431, "ref_ms": 4.02665, "cand_ms": 1.7783, "endpoint": "endpoint-04", "r_wall_s": 42.96, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 9060, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00142", "task": "L2_069", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "show_all", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_069|claude-opus-5-5|1", "design": "f1", "success": true, "score": 1.2512, "status": "ok", "correct": true, "speedup": 1.2512, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 124.8, "calls": 2, "tokens_in": 16023, "tokens_out": 3771, "ctx_chars": 12471, "ctx_sha": "4ddcc848d0210ca9", "late_delivered": null, "t_end": 1791553240.3842413, "backend": "triton", "own_frac": 1.0, "ref_ms": 3.65999, "cand_ms": 2.9252, "endpoint": "endpoint-26", "r_wall_s": 66.63, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 2713, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00133", "task": "L1_065", "split": "calib", "reader": "claude-opus-5-5", "arm": "none", "label": "none", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L1_065|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.9412, "status": "ok", "correct": true, "speedup": 1.9412, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 186.9, "calls": 2, "tokens_in": 7445, "tokens_out": 10535, "ctx_chars": 0, "ctx_sha": "e3b0c44298fc1c14", "late_delivered": null, "t_end": 1791553240.5343506, "backend": "cuda", "own_frac": 0.9998, "ref_ms": 43.88425, "cand_ms": 22.60694, "endpoint": "endpoint-10", "r_wall_s": 69.58, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 7742, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00146", "task": "L2_046", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_approach", "label": "no_approach", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_046|claude-opus-5-5|1", "design": "f1", "success": true, "score": 1.4022, "status": "ok", "correct": true, "speedup": 1.4022, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 113.3, "calls": 2, "tokens_in": 7280, "tokens_out": 3260, "ctx_chars": 3080, "ctx_sha": "7364d33fba159fed", "late_delivered": null, "t_end": 1791553247.4028628, "backend": "triton", "own_frac": 0.0611, "ref_ms": 19.2769, "cand_ms": 13.74729, "endpoint": "endpoint-01", "r_wall_s": 53.64, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 2279, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00143", "task": "L2_085", "split": "calib", "reader": "claude-opus-5-5", "arm": "frontier_gated", "label": "frontier_gated", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_085|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.6017, "status": "ok", "correct": true, "speedup": 1.6017, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 120.3, "calls": 2, "tokens_in": 12745, "tokens_out": 5345, "ctx_chars": 7376, "ctx_sha": "cc69cebe7b058084", "late_delivered": null, "t_end": 1791553247.434878, "backend": "triton", "own_frac": 0.2855, "ref_ms": 4.75791, "cand_ms": 2.97054, "endpoint": "endpoint-08", "r_wall_s": 21.53, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 3383, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00144", "task": "L2_069", "split": "calib", "reader": "claude-opus-5-5", "arm": "frontier_gated", "label": "frontier_gated", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_069|claude-opus-5-5|1", "design": "f1", "success": false, "score": 0.0, "status": "runtime_error", "correct": false, "speedup": 0.0, "infra": false, "infra_reruns": 0, "error": "OutOfMemoryError: CUDA out of memory. Tried to allocate 498.00 MiB. GPU 0 has a total capacity of 39.49 GiB of which 372.81 MiB is free. Including non-PyTorch memory, this process has 814.00 MiB memory in use. Of the allocated memory 320.02 MiB is allocated by PyTorch, and 1.98 MiB is reserved by Py", "wall_s": 143.8, "calls": 2, "tokens_in": 12755, "tokens_out": 9790, "ctx_chars": 7376, "ctx_sha": "cc69cebe7b058084", "late_delivered": null, "t_end": 1791553272.2198322, "backend": "cuda", "own_frac": null, "ref_ms": null, "cand_ms": null, "endpoint": "endpoint-13", "r_wall_s": 16.38, "flags": null, "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 7361, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00135", "task": "L3_038", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "late_show_all", "delivery": "late", "rep": 1, "p_arm": 0.111111, "block": "L3_038|claude-opus-5-5|1", "design": "f1", "success": true, "score": 3.0981, "status": "ok", "correct": true, "speedup": 3.0981, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 197.1, "calls": 3, "tokens_in": 23551, "tokens_out": 14744, "ctx_chars": 8224, "ctx_sha": "71fd1a97cbd02e3b", "late_delivered": true, "t_end": 1791553273.7835445, "backend": "cuda", "own_frac": 0.8731, "ref_ms": 80.70239, "cand_ms": 26.04909, "endpoint": "endpoint-08", "r_wall_s": 20.61, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 8064, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00140", "task": "L2_018", "split": "calib", "reader": "claude-opus-5-5", "arm": "reader_relative", "label": "reader_relative", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_018|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.0102, "status": "ok", "correct": true, "speedup": 1.0102, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 168.8, "calls": 4, "tokens_in": 35558, "tokens_out": 11010, "ctx_chars": 10525, "ctx_sha": "ea536f3d179af7b3", "late_delivered": null, "t_end": 1791553277.008162, "backend": "triton", "own_frac": 0.0034, "ref_ms": 8.03108, "cand_ms": 7.94965, "endpoint": "endpoint-01", "r_wall_s": 17.39, "flags": [], "submit_turn": 3, "turns": 4, "tech_src": "triton", "tok_out_first": 2712, "stops": ["end_turn", "end_turn", "end_turn", "end_turn"], "first_check": "incorrect"} {"ep": "f1-00122", "task": "L3_039", "split": "calib", "reader": "claude-opus-5-5", "arm": "frontier_gated", "label": "frontier_gated", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L3_039|claude-opus-5-5|1", "design": "f1", "success": true, "score": 6.7302, "status": "ok", "correct": true, "speedup": 6.7302, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 314.1, "calls": 3, "tokens_in": 21356, "tokens_out": 19997, "ctx_chars": 2260, "ctx_sha": "9b6af621f8f8a88b", "late_delivered": null, "t_end": 1791553281.2145119, "backend": "cuda", "own_frac": 0.9848, "ref_ms": 31.96247, "cand_ms": 4.74914, "endpoint": "endpoint-18", "r_wall_s": 50.38, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 12039, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00151", "task": "L2_073", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "show_all", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_073|claude-opus-5-5|0", "design": "f1", "success": false, "score": 0.0, "status": "runtime_error", "correct": false, "speedup": 0.0, "infra": false, "infra_reruns": 0, "error": "AcceleratorError: CUDA error: out of memory\nSearch for `cudaErrorMemoryAllocation' in https://docs.nvidia.com/cuda/cuda-runtime-api/group__CUDART__TYPES.html for more information.\nCUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.\n", "wall_s": 97.5, "calls": 3, "tokens_in": 22857, "tokens_out": 6862, "ctx_chars": 9590, "ctx_sha": "5bdedd614e688703", "late_delivered": null, "t_end": 1791553282.0162039, "backend": "triton", "own_frac": null, "ref_ms": null, "cand_ms": null, "endpoint": "endpoint-13", "r_wall_s": 7.26, "flags": null, "submit_turn": 2, "turns": 3, "tech_src": "triton", "tok_out_first": 3577, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "forbidden"} {"ep": "f1-00149", "task": "L2_067", "split": "calib", "reader": "claude-opus-5-5", "arm": "none", "label": "none", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_067|claude-opus-5-5|0", "design": "f1", "success": true, "score": 2.1319, "status": "ok", "correct": true, "speedup": 2.1319, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 122.7, "calls": 2, "tokens_in": 6357, "tokens_out": 9158, "ctx_chars": 0, "ctx_sha": "e3b0c44298fc1c14", "late_delivered": null, "t_end": 1791553288.5727358, "backend": "cuda", "own_frac": 0.9997, "ref_ms": 12.87111, "cand_ms": 6.03747, "endpoint": "endpoint-23", "r_wall_s": 28.83, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 6643, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00141", "task": "L1_083", "split": "calib", "reader": "claude-opus-5-5", "arm": "frontier_gated", "label": "frontier_gated", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L1_083|claude-opus-5-5|1", "design": "f1", "success": true, "score": 3.1263, "status": "ok", "correct": true, "speedup": 3.1263, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 179.9, "calls": 3, "tokens_in": 19644, "tokens_out": 9142, "ctx_chars": 4252, "ctx_sha": "7621aa423896afd1", "late_delivered": null, "t_end": 1791553293.1638076, "backend": "cuda", "own_frac": 1.0, "ref_ms": 2.92309, "cand_ms": 0.93499, "endpoint": "endpoint-05", "r_wall_s": 60.49, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 4001, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "runtime_error"} {"ep": "f1-00147", "task": "L2_069", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "late_show_all", "delivery": "late", "rep": 1, "p_arm": 0.111111, "block": "L2_069|claude-opus-5-5|1", "design": "f1", "success": false, "score": 0.0, "status": "runtime_error", "correct": false, "speedup": 0.0, "infra": false, "infra_reruns": 0, "error": "OutOfMemoryError: CUDA out of memory. Tried to allocate 498.00 MiB. GPU 0 has a total capacity of 39.49 GiB of which 372.81 MiB is free. Including non-PyTorch memory, this process has 814.00 MiB memory in use. Of the allocated memory 320.02 MiB is allocated by PyTorch, and 1.98 MiB is reserved by Py", "wall_s": 157.1, "calls": 3, "tokens_in": 24322, "tokens_out": 10134, "ctx_chars": 12471, "ctx_sha": "4ddcc848d0210ca9", "late_delivered": true, "t_end": 1791553299.6387339, "backend": "cuda", "own_frac": null, "ref_ms": null, "cand_ms": null, "endpoint": "endpoint-12", "r_wall_s": 16.05, "flags": null, "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 5112, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00150", "task": "L1_065", "split": "calib", "reader": "claude-opus-5-5", "arm": "lessons_only", "label": "lessons_only", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L1_065|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.6601, "status": "ok", "correct": true, "speedup": 1.6601, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 138.4, "calls": 2, "tokens_in": 7876, "tokens_out": 5585, "ctx_chars": 1982, "ctx_sha": "54e16caf9bef9eec", "late_delivered": null, "t_end": 1791553309.3670602, "backend": "triton", "own_frac": 0.9998, "ref_ms": 44.65046, "cand_ms": 26.89626, "endpoint": "endpoint-24", "r_wall_s": 69.9, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 3774, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00158", "task": "L1_065", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "show_all", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L1_065|claude-opus-5-5|1", "design": "f1", "success": true, "score": 2.2023, "status": "ok", "correct": true, "speedup": 2.2023, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 90.5, "calls": 2, "tokens_in": 15098, "tokens_out": 2264, "ctx_chars": 11175, "ctx_sha": "91e4a03c424736b2", "late_delivered": null, "t_end": 1791553310.040144, "backend": "triton", "own_frac": 0.0002, "ref_ms": 44.45165, "cand_ms": 20.18436, "endpoint": "endpoint-07", "r_wall_s": 57.1, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 1210, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00145", "task": "L2_027", "split": "calib", "reader": "claude-opus-5-5", "arm": "none", "label": "none", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_027|claude-opus-5-5|1", "design": "f1", "success": true, "score": 4.0069, "status": "ok", "correct": true, "speedup": 4.0069, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 183.1, "calls": 2, "tokens_in": 7547, "tokens_out": 14182, "ctx_chars": 0, "ctx_sha": "e3b0c44298fc1c14", "late_delivered": null, "t_end": 1791553312.9455485, "backend": "cuda", "own_frac": 0.9992, "ref_ms": 20.03859, "cand_ms": 5.00105, "endpoint": "endpoint-26", "r_wall_s": 27.56, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 10811, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00155", "task": "L2_025", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "late_show_all", "delivery": "late", "rep": 1, "p_arm": 0.111111, "block": "L2_025|claude-opus-5-5|1", "design": "f1", "success": true, "score": 1.8272, "status": "ok", "correct": true, "speedup": 1.8272, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 135.5, "calls": 2, "tokens_in": 10464, "tokens_out": 9411, "ctx_chars": 10087, "ctx_sha": "fc8a2930809fd512", "late_delivered": true, "t_end": 1791553331.1397264, "backend": "cuda", "own_frac": 1.0, "ref_ms": 19.23111, "cand_ms": 10.52514, "endpoint": "endpoint-25", "r_wall_s": 25.31, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 6690, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00137", "task": "L3_036", "split": "calib", "reader": "claude-opus-5-5", "arm": "reader_relative", "label": "reader_relative", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L3_036|claude-opus-5-5|0", "design": "f1", "success": true, "score": 6.2107, "status": "ok", "correct": true, "speedup": 6.2107, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 241.7, "calls": 3, "tokens_in": 29852, "tokens_out": 21600, "ctx_chars": 8756, "ctx_sha": "26dc0616aa00d951", "late_delivered": null, "t_end": 1791553332.5170414, "backend": "cuda", "own_frac": 0.9766, "ref_ms": 31.55869, "cand_ms": 5.08136, "endpoint": "endpoint-18", "r_wall_s": 27.6, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 13457, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00165", "task": "L2_074", "split": "calib", "reader": "claude-opus-5-5", "arm": "frontier_gated", "label": "frontier_gated", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_074|claude-opus-5-5|1", "design": "f1", "success": true, "score": 1.3715, "status": "ok", "correct": true, "speedup": 1.3715, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 66.3, "calls": 2, "tokens_in": 8035, "tokens_out": 3954, "ctx_chars": 3256, "ctx_sha": "8fa366c59fab0c7f", "late_delivered": null, "t_end": 1791553343.281021, "backend": "triton", "own_frac": 0.2924, "ref_ms": 13.3616, "cand_ms": 9.74263, "endpoint": "endpoint-04", "r_wall_s": 18.58, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 2593, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00156", "task": "L2_008", "split": "calib", "reader": "claude-opus-5-5", "arm": "lessons_only", "label": "lessons_only", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_008|claude-opus-5-5|0", "design": "f1", "success": true, "score": 4.4654, "status": "ok", "correct": true, "speedup": 4.4654, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 156.5, "calls": 2, "tokens_in": 9079, "tokens_out": 12783, "ctx_chars": 1982, "ctx_sha": "54e16caf9bef9eec", "late_delivered": null, "t_end": 1791553360.9556446, "backend": "cuda", "own_frac": 0.9993, "ref_ms": 12.70782, "cand_ms": 2.84584, "endpoint": "endpoint-10", "r_wall_s": 27.24, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 9596, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00160", "task": "L1_078", "split": "calib", "reader": "claude-opus-5-5", "arm": "reader_relative", "label": "reader_relative", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L1_078|claude-opus-5-5|1", "design": "f1", "success": false, "score": 0.0, "status": "runtime_error", "correct": false, "speedup": 0.0, "infra": false, "infra_reruns": 0, "error": "OutOfMemoryError: CUDA out of memory. Tried to allocate 1024.00 MiB. GPU 0 has a total capacity of 39.49 GiB of which 108.81 MiB is free. Including non-PyTorch memory, this process has 1.16 GiB memory in use. Of the allocated memory 768.08 MiB is allocated by PyTorch, and 1.92 MiB is reserved by PyT", "wall_s": 121.9, "calls": 2, "tokens_in": 17068, "tokens_out": 7324, "ctx_chars": 12197, "ctx_sha": "d2edcc521b94d748", "late_delivered": null, "t_end": 1791553362.4073825, "backend": "triton", "own_frac": null, "ref_ms": null, "cand_ms": null, "endpoint": "endpoint-13", "r_wall_s": 7.21, "flags": null, "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 5537, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00148", "task": "L1_083", "split": "calib", "reader": "claude-opus-5-5", "arm": "none", "label": "none", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L1_083|claude-opus-5-5|1", "design": "f1", "success": true, "score": 2.6237, "status": "ok", "correct": true, "speedup": 2.6237, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 216.1, "calls": 3, "tokens_in": 15953, "tokens_out": 10055, "ctx_chars": 0, "ctx_sha": "e3b0c44298fc1c14", "late_delivered": null, "t_end": 1791553365.930898, "backend": "cuda", "own_frac": 1.0, "ref_ms": 2.37249, "cand_ms": 0.90426, "endpoint": "endpoint-01", "r_wall_s": 73.52, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 3876, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "runtime_error"} {"ep": "f1-00164", "task": "L1_062", "split": "calib", "reader": "claude-opus-5-5", "arm": "lessons_only", "label": "lessons_only", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L1_062|claude-opus-5-5|1", "design": "f1", "success": false, "score": 0.0, "status": "runtime_error", "correct": false, "speedup": 0.0, "infra": false, "infra_reruns": 0, "error": "OutOfMemoryError: CUDA out of memory. Tried to allocate 512.00 MiB. GPU 0 has a total capacity of 39.49 GiB of which 28.81 MiB is free. Including non-PyTorch memory, this process has 1.48 GiB memory in use. Of the allocated memory 1.00 GiB is allocated by PyTorch, and 1.65 MiB is reserved by PyTorch", "wall_s": 97.8, "calls": 2, "tokens_in": 8025, "tokens_out": 7466, "ctx_chars": 1982, "ctx_sha": "54e16caf9bef9eec", "late_delivered": null, "t_end": 1791553371.5803995, "backend": "triton", "own_frac": null, "ref_ms": null, "cand_ms": null, "endpoint": "endpoint-12", "r_wall_s": 6.6, "flags": null, "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 5462, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00168", "task": "L3_046", "split": "calib", "reader": "claude-opus-5-5", "arm": "frontier_gated", "label": "frontier_gated", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L3_046|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.4015, "status": "ok", "correct": true, "speedup": 1.4015, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 90.9, "calls": 2, "tokens_in": 11236, "tokens_out": 6418, "ctx_chars": 4592, "ctx_sha": "9ac9f3e444d8f2ff", "late_delivered": null, "t_end": 1791553379.4589717, "backend": "triton", "own_frac": 0.0843, "ref_ms": 4.0022, "cand_ms": 2.85563, "endpoint": "endpoint-11", "r_wall_s": 22.86, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 4786, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00161", "task": "L2_018", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "show_all", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_018|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.0118, "status": "ok", "correct": true, "speedup": 1.0118, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 135.6, "calls": 4, "tokens_in": 32215, "tokens_out": 8556, "ctx_chars": 9898, "ctx_sha": "eb48d1e5953cd4b8", "late_delivered": null, "t_end": 1791553383.0115683, "backend": "triton", "own_frac": 0.003, "ref_ms": 8.09697, "cand_ms": 8.00263, "endpoint": "endpoint-24", "r_wall_s": 13.61, "flags": [], "submit_turn": 3, "turns": 4, "tech_src": "triton", "tok_out_first": 1697, "stops": ["end_turn", "end_turn", "end_turn", "end_turn"], "first_check": "incorrect"} {"ep": "f1-00152", "task": "L1_078", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "late_show_all", "delivery": "late", "rep": 1, "p_arm": 0.111111, "block": "L1_078|claude-opus-5-5|1", "design": "f1", "success": true, "score": 3.1197, "status": "ok", "correct": true, "speedup": 3.1197, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 195.5, "calls": 2, "tokens_in": 12531, "tokens_out": 10669, "ctx_chars": 11582, "ctx_sha": "a74f6646d7158b72", "late_delivered": true, "t_end": 1791553385.4912245, "backend": "cuda", "own_frac": 1.0, "ref_ms": 35.51568, "cand_ms": 11.38434, "endpoint": "endpoint-19", "r_wall_s": 82.76, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 7405, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00154", "task": "L3_042", "split": "calib", "reader": "claude-opus-5-5", "arm": "none", "label": "none", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L3_042|claude-opus-5-5|0", "design": "f1", "success": true, "score": 3.0195, "status": "ok", "correct": true, "speedup": 3.0195, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 206.1, "calls": 3, "tokens_in": 16834, "tokens_out": 14413, "ctx_chars": 0, "ctx_sha": "e3b0c44298fc1c14", "late_delivered": null, "t_end": 1791553397.0375466, "backend": "cuda", "own_frac": 0.8656, "ref_ms": 59.52208, "cand_ms": 19.71231, "endpoint": "endpoint-16", "r_wall_s": 45.61, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 7742, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00159", "task": "L2_027", "split": "calib", "reader": "claude-opus-5-5", "arm": "fg_target", "label": "fg_target", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_027|claude-opus-5-5|1", "design": "f1", "success": true, "score": 3.9849, "status": "ok", "correct": true, "speedup": 3.9849, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 159.8, "calls": 2, "tokens_in": 13860, "tokens_out": 13810, "ctx_chars": 7376, "ctx_sha": "cc69cebe7b058084", "late_delivered": null, "t_end": 1791553400.1391952, "backend": "cuda", "own_frac": 0.9996, "ref_ms": 20.00824, "cand_ms": 5.02104, "endpoint": "endpoint-05", "r_wall_s": 18.97, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 10585, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00157", "task": "L1_071", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "late_show_all", "delivery": "late", "rep": 1, "p_arm": 0.111111, "block": "L1_071|claude-opus-5-5|1", "design": "f1", "success": false, "score": 0.5636, "status": "ok", "correct": true, "speedup": 0.5636, "infra": false, "infra_reruns": 0, "error": "ok", "wall_s": 196.3, "calls": 2, "tokens_in": 11920, "tokens_out": 9739, "ctx_chars": 11175, "ctx_sha": "91e4a03c424736b2", "late_delivered": true, "t_end": 1791553405.1740568, "backend": "cuda", "own_frac": 1.0, "ref_ms": 3.34116, "cand_ms": 5.92808, "endpoint": "endpoint-17", "r_wall_s": 72.54, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 6578, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00173", "task": "L1_083", "split": "calib", "reader": "claude-opus-5-5", "arm": "fg_target", "label": "fg_target", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L1_083|claude-opus-5-5|1", "design": "f1", "success": true, "score": 3.2343, "status": "ok", "correct": true, "speedup": 3.2343, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 104.2, "calls": 2, "tokens_in": 7742, "tokens_out": 3331, "ctx_chars": 3033, "ctx_sha": "1962fe92a5a60723", "late_delivered": null, "t_end": 1791553417.1367745, "backend": "triton", "own_frac": 1.0, "ref_ms": 2.93003, "cand_ms": 0.90594, "endpoint": "endpoint-26", "r_wall_s": 53.45, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 2041, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00162", "task": "L2_067", "split": "calib", "reader": "claude-opus-5-5", "arm": "lessons_only", "label": "lessons_only", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_067|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.8249, "status": "ok", "correct": true, "speedup": 1.8249, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 174.4, "calls": 3, "tokens_in": 15831, "tokens_out": 11941, "ctx_chars": 1982, "ctx_sha": "54e16caf9bef9eec", "late_delivered": null, "t_end": 1791553421.8566215, "backend": "cuda", "own_frac": 0.9997, "ref_ms": 12.82065, "cand_ms": 7.02534, "endpoint": "endpoint-11", "r_wall_s": 36.02, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 6842, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00174", "task": "L2_011", "split": "calib", "reader": "claude-opus-5-5", "arm": "reader_relative", "label": "reader_relative", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_011|claude-opus-5-5|1", "design": "f1", "success": true, "score": 4.0912, "status": "ok", "correct": true, "speedup": 4.0912, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 92.9, "calls": 2, "tokens_in": 18338, "tokens_out": 7307, "ctx_chars": 12980, "ctx_sha": "7f54a0d591220826", "late_delivered": null, "t_end": 1791553424.082688, "backend": "triton", "own_frac": 0.0804, "ref_ms": 27.92145, "cand_ms": 6.82475, "endpoint": "endpoint-16", "r_wall_s": 19.75, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 4804, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00170", "task": "L2_011", "split": "calib", "reader": "claude-opus-5-5", "arm": "lessons_only", "label": "lessons_only", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_011|claude-opus-5-5|1", "design": "f1", "success": true, "score": 1.0805, "status": "ok", "correct": true, "speedup": 1.0805, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 136.2, "calls": 2, "tokens_in": 8099, "tokens_out": 8709, "ctx_chars": 1982, "ctx_sha": "54e16caf9bef9eec", "late_delivered": null, "t_end": 1791553435.883229, "backend": "triton", "own_frac": 0.0204, "ref_ms": 27.78938, "cand_ms": 25.71857, "endpoint": "endpoint-18", "r_wall_s": 21.76, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 6442, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00166", "task": "L2_067", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "show_all", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_067|claude-opus-5-5|0", "design": "f1", "success": false, "score": 0.0832, "status": "ok", "correct": true, "speedup": 0.0832, "infra": false, "infra_reruns": 0, "error": "ok", "wall_s": 172.9, "calls": 2, "tokens_in": 14705, "tokens_out": 5339, "ctx_chars": 11023, "ctx_sha": "04625c0d6e6734eb", "late_delivered": null, "t_end": 1791553454.1034253, "backend": "triton", "own_frac": 1.0, "ref_ms": 12.88316, "cand_ms": 154.84494, "endpoint": "endpoint-25", "r_wall_s": 57.69, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 4190, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00177", "task": "L1_071", "split": "calib", "reader": "claude-opus-5-5", "arm": "reader_relative", "label": "reader_relative", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L1_071|claude-opus-5-5|1", "design": "f1", "success": false, "score": 0.7343, "status": "ok", "correct": true, "speedup": 0.7343, "infra": false, "infra_reruns": 0, "error": "ok", "wall_s": 93.6, "calls": 2, "tokens_in": 15703, "tokens_out": 2581, "ctx_chars": 11807, "ctx_sha": "04bab9b1ff7e3fde", "late_delivered": null, "t_end": 1791553454.5891562, "backend": "triton", "own_frac": 0.0008, "ref_ms": 3.30829, "cand_ms": 4.50534, "endpoint": "endpoint-10", "r_wall_s": 53.57, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 1553, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00175", "task": "L2_025", "split": "calib", "reader": "claude-opus-5-5", "arm": "lessons_only", "label": "lessons_only", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_025|claude-opus-5-5|0", "design": "f1", "success": true, "score": 2.1175, "status": "ok", "correct": true, "speedup": 2.1175, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 130.7, "calls": 2, "tokens_in": 7382, "tokens_out": 9624, "ctx_chars": 1982, "ctx_sha": "54e16caf9bef9eec", "late_delivered": null, "t_end": 1791553463.2223456, "backend": "cuda", "own_frac": 1.0, "ref_ms": 19.14481, "cand_ms": 9.04127, "endpoint": "endpoint-05", "r_wall_s": 23.15, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 7545, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00139", "task": "L1_085", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_approach", "label": "no_approach", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L1_085|claude-opus-5-5|0", "design": "f1", "success": true, "score": 3.3575, "status": "ok", "correct": true, "speedup": 3.3575, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 367.6, "calls": 3, "tokens_in": 21556, "tokens_out": 15060, "ctx_chars": 3080, "ctx_sha": "7364d33fba159fed", "late_delivered": null, "t_end": 1791553464.4427888, "backend": "cuda", "own_frac": 1.0, "ref_ms": 3.73727, "cand_ms": 1.11313, "endpoint": "endpoint-23", "r_wall_s": 97.47, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 8165, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00169", "task": "L3_046", "split": "calib", "reader": "claude-opus-5-5", "arm": "frontier_gated", "label": "frontier_gated", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L3_046|claude-opus-5-5|1", "design": "f1", "success": true, "score": 2.4207, "status": "ok", "correct": true, "speedup": 2.4207, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 174.0, "calls": 2, "tokens_in": 11827, "tokens_out": 9096, "ctx_chars": 4592, "ctx_sha": "9ac9f3e444d8f2ff", "late_delivered": null, "t_end": 1791553467.134976, "backend": "triton", "own_frac": 0.4022, "ref_ms": 4.01349, "cand_ms": 1.65798, "endpoint": "endpoint-19", "r_wall_s": 50.64, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 7029, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00181", "task": "L3_038", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_numbers", "label": "no_numbers", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L3_038|claude-opus-5-5|0", "design": "f1", "success": false, "score": 0.8013, "status": "ok", "correct": true, "speedup": 0.8013, "infra": false, "infra_reruns": 0, "error": "ok", "wall_s": 98.8, "calls": 2, "tokens_in": 13016, "tokens_out": 5555, "ctx_chars": 8204, "ctx_sha": "83777d4de17c3440", "late_delivered": null, "t_end": 1791553478.2415254, "backend": "triton", "own_frac": 0.0, "ref_ms": 79.11865, "cand_ms": 98.73709, "endpoint": "endpoint-08", "r_wall_s": 15.35, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 4086, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00171", "task": "L1_078", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_approach", "label": "no_approach", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L1_078|claude-opus-5-5|1", "design": "f1", "success": true, "score": 2.2696, "status": "ok", "correct": true, "speedup": 2.2696, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 180.6, "calls": 2, "tokens_in": 8460, "tokens_out": 6203, "ctx_chars": 3030, "ctx_sha": "e6667ea63ccdb086", "late_delivered": null, "t_end": 1791553490.0138907, "backend": "triton", "own_frac": 1.0, "ref_ms": 36.03757, "cand_ms": 15.87821, "endpoint": "endpoint-24", "r_wall_s": 82.12, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 4592, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00188", "task": "L2_082", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "show_all", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_082|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.933, "status": "ok", "correct": true, "speedup": 1.933, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 72.1, "calls": 2, "tokens_in": 14618, "tokens_out": 4019, "ctx_chars": 10743, "ctx_sha": "30d3b01036fcb086", "late_delivered": null, "t_end": 1791553493.96131, "backend": "triton", "own_frac": 0.1507, "ref_ms": 19.41127, "cand_ms": 10.04183, "endpoint": "endpoint-11", "r_wall_s": 23.33, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 2885, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00182", "task": "L1_083", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_approach", "label": "no_approach", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L1_083|claude-opus-5-5|0", "design": "f1", "success": true, "score": 2.6152, "status": "ok", "correct": true, "speedup": 2.6152, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 111.7, "calls": 2, "tokens_in": 7830, "tokens_out": 3470, "ctx_chars": 3080, "ctx_sha": "7364d33fba159fed", "late_delivered": null, "t_end": 1791553494.71708, "backend": "triton", "own_frac": 1.0, "ref_ms": 2.38104, "cand_ms": 0.91045, "endpoint": "endpoint-16", "r_wall_s": 58.7, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 2194, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00183", "task": "L1_072", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "show_all", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L1_072|claude-opus-5-5|1", "design": "f1", "success": true, "score": 2.1771, "status": "ok", "correct": true, "speedup": 2.1771, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 125.5, "calls": 2, "tokens_in": 16813, "tokens_out": 10281, "ctx_chars": 11175, "ctx_sha": "91e4a03c424736b2", "late_delivered": null, "t_end": 1791553510.9583466, "backend": "triton", "own_frac": 1.0, "ref_ms": 4.38463, "cand_ms": 2.01399, "endpoint": "endpoint-23", "r_wall_s": 24.64, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 8101, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00163", "task": "L3_049", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_numbers", "label": "no_numbers", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L3_049|claude-opus-5-5|1", "design": "f1", "success": false, "score": 0.0, "status": "runtime_error", "correct": false, "speedup": 0.0, "infra": false, "infra_reruns": 0, "error": "OutOfMemoryError: CUDA out of memory. Tried to allocate 512.00 MiB. GPU 0 has a total capacity of 39.49 GiB of which 359.94 MiB is free. Including non-PyTorch memory, this process has 948.00 MiB memory in use. Of the allocated memory 520.00 MiB is allocated by PyTorch, and 12.00 MiB is reserved by P", "wall_s": 245.6, "calls": 4, "tokens_in": 31792, "tokens_out": 16047, "ctx_chars": 2425, "ctx_sha": "5c5e8811718de973", "late_delivered": null, "t_end": 1791553517.8029149, "backend": "cuda", "own_frac": null, "ref_ms": null, "cand_ms": null, "endpoint": "endpoint-08", "r_wall_s": 14.86, "flags": null, "submit_turn": 3, "turns": 4, "tech_src": "cuda", "tok_out_first": 5644, "stops": ["end_turn", "end_turn", "end_turn", "end_turn"], "first_check": "incorrect"} {"ep": "f1-00190", "task": "L2_046", "split": "calib", "reader": "claude-opus-5-5", "arm": "frontier_gated", "label": "frontier_gated", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_046|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.3994, "status": "ok", "correct": true, "speedup": 1.3994, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 82.0, "calls": 2, "tokens_in": 7757, "tokens_out": 2971, "ctx_chars": 3639, "ctx_sha": "fc6e42d43dfb4e75", "late_delivered": null, "t_end": 1791553517.9296656, "backend": "triton", "own_frac": 0.0598, "ref_ms": 19.42684, "cand_ms": 13.88205, "endpoint": "endpoint-09", "r_wall_s": 36.05, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 2040, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00186", "task": "L1_085", "split": "calib", "reader": "claude-opus-5-5", "arm": "frontier_gated", "label": "frontier_gated", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L1_085|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.3016, "status": "ok", "correct": true, "speedup": 1.3016, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 118.1, "calls": 2, "tokens_in": 13310, "tokens_out": 3879, "ctx_chars": 7816, "ctx_sha": "79f0910e887d8bbe", "late_delivered": null, "t_end": 1791553523.2923763, "backend": "triton", "own_frac": 1.0, "ref_ms": 3.75227, "cand_ms": 2.88276, "endpoint": "endpoint-10", "r_wall_s": 64.97, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 2303, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00191", "task": "L1_071", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_numbers", "label": "no_numbers", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L1_071|claude-opus-5-5|0", "design": "f1", "success": false, "score": 0.7301, "status": "ok", "correct": true, "speedup": 0.7301, "infra": false, "infra_reruns": 0, "error": "ok", "wall_s": 110.6, "calls": 2, "tokens_in": 14967, "tokens_out": 2186, "ctx_chars": 11170, "ctx_sha": "ee0e99e621c34bdc", "late_delivered": null, "t_end": 1791553564.6895225, "backend": "triton", "own_frac": 0.0006, "ref_ms": 3.29894, "cand_ms": 4.51852, "endpoint": "endpoint-17", "r_wall_s": 65.86, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 1152, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00195", "task": "L3_008", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "show_all", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L3_008|claude-opus-5-5|1", "design": "f1", "success": true, "score": 1.3097, "status": "ok", "correct": true, "speedup": 1.3097, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 101.8, "calls": 2, "tokens_in": 12257, "tokens_out": 7231, "ctx_chars": 6513, "ctx_sha": "1aa698d3a9589f25", "late_delivered": null, "t_end": 1791553568.9489772, "backend": "triton", "own_frac": 0.0591, "ref_ms": 5.25084, "cand_ms": 4.00931, "endpoint": "endpoint-12", "r_wall_s": 24.88, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 4845, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00192", "task": "L1_050", "split": "calib", "reader": "claude-opus-5-5", "arm": "lessons_only", "label": "lessons_only", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L1_050|claude-opus-5-5|1", "design": "f1", "success": true, "score": 1.7729, "status": "ok", "correct": true, "speedup": 1.7729, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 115.9, "calls": 2, "tokens_in": 6496, "tokens_out": 6315, "ctx_chars": 1982, "ctx_sha": "54e16caf9bef9eec", "late_delivered": null, "t_end": 1791553570.5304694, "backend": "triton", "own_frac": 0.998, "ref_ms": 5.28412, "cand_ms": 2.98053, "endpoint": "endpoint-11", "r_wall_s": 39.83, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 4768, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00176", "task": "L2_024", "split": "calib", "reader": "claude-opus-5-5", "arm": "frontier_gated", "label": "frontier_gated", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_024|claude-opus-5-5|1", "design": "f1", "success": false, "score": 0.2898, "status": "ok", "correct": true, "speedup": 0.2898, "infra": false, "infra_reruns": 0, "error": "ok", "wall_s": 240.2, "calls": 3, "tokens_in": 16118, "tokens_out": 7859, "ctx_chars": 3639, "ctx_sha": "fc6e42d43dfb4e75", "late_delivered": null, "t_end": 1791553583.457427, "backend": "cuda", "own_frac": 0.9994, "ref_ms": 2.3772, "cand_ms": 8.20274, "endpoint": "endpoint-09", "r_wall_s": 65.18, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 4043, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00189", "task": "L2_067", "split": "calib", "reader": "claude-opus-5-5", "arm": "none", "label": "none", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_067|claude-opus-5-5|1", "design": "f1", "success": true, "score": 2.1352, "status": "ok", "correct": true, "speedup": 2.1352, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 162.0, "calls": 3, "tokens_in": 14071, "tokens_out": 12097, "ctx_chars": 0, "ctx_sha": "e3b0c44298fc1c14", "late_delivered": null, "t_end": 1791553586.0432508, "backend": "cuda", "own_frac": 0.9997, "ref_ms": 12.94187, "cand_ms": 6.06117, "endpoint": "endpoint-16", "r_wall_s": 23.72, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 6604, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00178", "task": "L1_080", "split": "calib", "reader": "claude-opus-5-5", "arm": "none", "label": "none", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L1_080|claude-opus-5-5|1", "design": "f1", "success": false, "score": 0.9315, "status": "ok", "correct": true, "speedup": 0.9315, "infra": false, "infra_reruns": 0, "error": "ok", "wall_s": 225.7, "calls": 2, "tokens_in": 7559, "tokens_out": 11790, "ctx_chars": 0, "ctx_sha": "e3b0c44298fc1c14", "late_delivered": null, "t_end": 1791553588.0798702, "backend": "cuda", "own_frac": 0.0002, "ref_ms": 20.23572, "cand_ms": 21.72284, "endpoint": "endpoint-07", "r_wall_s": 100.7, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 8743, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00196", "task": "L2_069", "split": "calib", "reader": "claude-opus-5-5", "arm": "reader_relative", "label": "reader_relative", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_069|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.1435, "status": "ok", "correct": true, "speedup": 1.1435, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 133.4, "calls": 2, "tokens_in": 16551, "tokens_out": 3680, "ctx_chars": 13102, "ctx_sha": "e5df41577725cdc6", "late_delivered": null, "t_end": 1791553611.6598938, "backend": "triton", "own_frac": 1.0, "ref_ms": 3.63758, "cand_ms": 3.18107, "endpoint": "endpoint-25", "r_wall_s": 68.97, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 2534, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00185", "task": "L3_036", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_approach", "label": "no_approach", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L3_036|claude-opus-5-5|1", "design": "f1", "success": true, "score": 6.2405, "status": "ok", "correct": true, "speedup": 6.2405, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 213.3, "calls": 3, "tokens_in": 21250, "tokens_out": 19204, "ctx_chars": 3355, "ctx_sha": "cc270772e526c41f", "late_delivered": null, "t_end": 1791553613.4249547, "backend": "cuda", "own_frac": 0.9996, "ref_ms": 32.88197, "cand_ms": 5.26915, "endpoint": "endpoint-09", "r_wall_s": 19.93, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 11720, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00184", "task": "L2_069", "split": "calib", "reader": "claude-opus-5-5", "arm": "reader_relative", "label": "reader_relative", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_069|claude-opus-5-5|1", "design": "f1", "success": false, "score": 0.6477, "status": "ok", "correct": true, "speedup": 0.6477, "infra": false, "infra_reruns": 0, "error": "ok", "wall_s": 218.1, "calls": 3, "tokens_in": 30167, "tokens_out": 10418, "ctx_chars": 13102, "ctx_sha": "e5df41577725cdc6", "late_delivered": null, "t_end": 1791553615.178483, "backend": "cuda", "own_frac": 0.9991, "ref_ms": 3.64084, "cand_ms": 5.62162, "endpoint": "endpoint-19", "r_wall_s": 69.45, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 5610, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00194", "task": "L2_074", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_approach", "label": "no_approach", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_074|claude-opus-5-5|1", "design": "f1", "success": true, "score": 9.2137, "status": "ok", "correct": true, "speedup": 9.2137, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 154.7, "calls": 3, "tokens_in": 18629, "tokens_out": 12622, "ctx_chars": 3030, "ctx_sha": "e6667ea63ccdb086", "late_delivered": null, "t_end": 1791553619.1745305, "backend": "cuda", "own_frac": 1.0, "ref_ms": 5.92616, "cand_ms": 0.64319, "endpoint": "endpoint-05", "r_wall_s": 21.5, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 6995, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00179", "task": "L2_018", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "late_show_all", "delivery": "late", "rep": 0, "p_arm": 0.111111, "block": "L2_018|claude-opus-5-5|0", "design": "f1", "success": true, "score": 5.4199, "status": "ok", "correct": true, "speedup": 5.4199, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 277.6, "calls": 3, "tokens_in": 25756, "tokens_out": 25015, "ctx_chars": 9898, "ctx_sha": "eb48d1e5953cd4b8", "late_delivered": true, "t_end": 1791553643.5002553, "backend": "cuda", "own_frac": 1.0, "ref_ms": 7.98877, "cand_ms": 1.47397, "endpoint": "endpoint-05", "r_wall_s": 20.47, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 4501, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "incorrect"} {"ep": "f1-00172", "task": "L3_049", "split": "calib", "reader": "claude-opus-5-5", "arm": "fg_target", "label": "fg_target", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L3_049|claude-opus-5-5|0", "design": "f1", "success": true, "score": 10.0, "status": "ok", "correct": true, "speedup": 10.2708, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 349.2, "calls": 3, "tokens_in": 20269, "tokens_out": 13277, "ctx_chars": 2578, "ctx_sha": "7b4237bf96a5394d", "late_delivered": null, "t_end": 1791553659.2731915, "backend": "cuda", "own_frac": 0.7631, "ref_ms": 11.97237, "cand_ms": 1.16567, "endpoint": "endpoint-24", "r_wall_s": 59.39, "flags": ["excessive_speedup>10.0x: audit"], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 5245, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "incorrect"} {"ep": "f1-00204", "task": "L2_025", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "show_all", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_025|claude-opus-5-5|1", "design": "f1", "success": true, "score": 1.4515, "status": "ok", "correct": true, "speedup": 1.4515, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 100.7, "calls": 2, "tokens_in": 13320, "tokens_out": 3877, "ctx_chars": 10087, "ctx_sha": "fc8a2930809fd512", "late_delivered": null, "t_end": 1791553665.4083588, "backend": "triton", "own_frac": 0.9997, "ref_ms": 19.27744, "cand_ms": 13.28137, "endpoint": "endpoint-18", "r_wall_s": 20.22, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 2739, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00193", "task": "L2_067", "split": "calib", "reader": "claude-opus-5-5", "arm": "frontier_gated", "label": "frontier_gated", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_067|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.7642, "status": "ok", "correct": true, "speedup": 1.7642, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 204.3, "calls": 3, "tokens_in": 18473, "tokens_out": 14982, "ctx_chars": 3639, "ctx_sha": "fc6e42d43dfb4e75", "late_delivered": null, "t_end": 1791553667.4843593, "backend": "cuda", "own_frac": 0.9993, "ref_ms": 12.86779, "cand_ms": 7.29366, "endpoint": "endpoint-09", "r_wall_s": 20.12, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 9332, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00197", "task": "L2_071", "split": "calib", "reader": "claude-opus-5-5", "arm": "none", "label": "none", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_071|claude-opus-5-5|1", "design": "f1", "success": true, "score": 1.9712, "status": "ok", "correct": true, "speedup": 1.9712, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 177.9, "calls": 2, "tokens_in": 6097, "tokens_out": 7686, "ctx_chars": 0, "ctx_sha": "e3b0c44298fc1c14", "late_delivered": null, "t_end": 1791553667.9324028, "backend": "cuda", "own_frac": 1.0, "ref_ms": 3.68561, "cand_ms": 1.86972, "endpoint": "endpoint-07", "r_wall_s": 73.95, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 5370, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00203", "task": "L2_024", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "late_show_all", "delivery": "late", "rep": 1, "p_arm": 0.111111, "block": "L2_024|claude-opus-5-5|1", "design": "f1", "success": false, "score": 0.227, "status": "ok", "correct": true, "speedup": 0.227, "infra": false, "infra_reruns": 0, "error": "ok", "wall_s": 155.8, "calls": 3, "tokens_in": 21065, "tokens_out": 8487, "ctx_chars": 10177, "ctx_sha": "cf1072ae0bde8a1b", "late_delivered": true, "t_end": 1791553679.1176517, "backend": "cuda", "own_frac": 0.9995, "ref_ms": 2.37183, "cand_ms": 10.44938, "endpoint": "endpoint-05", "r_wall_s": 34.6, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 4171, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00205", "task": "L1_065", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_numbers", "label": "no_numbers", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L1_065|claude-opus-5-5|1", "design": "f1", "success": true, "score": 2.1903, "status": "ok", "correct": true, "speedup": 2.1903, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 111.2, "calls": 2, "tokens_in": 15161, "tokens_out": 2455, "ctx_chars": 11170, "ctx_sha": "ee0e99e621c34bdc", "late_delivered": null, "t_end": 1791553680.1563733, "backend": "triton", "own_frac": 0.0002, "ref_ms": 44.36536, "cand_ms": 20.25583, "endpoint": "endpoint-16", "r_wall_s": 69.96, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 1327, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00202", "task": "L2_008", "split": "calib", "reader": "claude-opus-5-5", "arm": "none", "label": "none", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_008|claude-opus-5-5|0", "design": "f1", "success": true, "score": 2.9397, "status": "ok", "correct": true, "speedup": 2.9397, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 164.8, "calls": 3, "tokens_in": 15658, "tokens_out": 13639, "ctx_chars": 0, "ctx_sha": "e3b0c44298fc1c14", "late_delivered": null, "t_end": 1791553682.7616756, "backend": "cuda", "own_frac": 0.9986, "ref_ms": 12.70484, "cand_ms": 4.32187, "endpoint": "endpoint-10", "r_wall_s": 21.87, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 7466, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00209", "task": "L1_065", "split": "calib", "reader": "claude-opus-5-5", "arm": "reader_relative", "label": "reader_relative", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L1_065|claude-opus-5-5|1", "design": "f1", "success": true, "score": 2.2295, "status": "ok", "correct": true, "speedup": 2.2295, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 112.2, "calls": 2, "tokens_in": 15433, "tokens_out": 2034, "ctx_chars": 11807, "ctx_sha": "04bab9b1ff7e3fde", "late_delivered": null, "t_end": 1791553700.276864, "backend": "triton", "own_frac": 0.0002, "ref_ms": 44.55776, "cand_ms": 19.98572, "endpoint": "endpoint-01", "r_wall_s": 60.22, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 1081, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00213", "task": "L3_046", "split": "calib", "reader": "claude-opus-5-5", "arm": "none", "label": "none", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L3_046|claude-opus-5-5|0", "design": "f1", "success": true, "score": 2.6052, "status": "ok", "correct": true, "speedup": 2.6052, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 108.8, "calls": 2, "tokens_in": 7968, "tokens_out": 7355, "ctx_chars": 0, "ctx_sha": "e3b0c44298fc1c14", "late_delivered": null, "t_end": 1791553727.9298115, "backend": "triton", "own_frac": 0.363, "ref_ms": 4.01395, "cand_ms": 1.54075, "endpoint": "endpoint-25", "r_wall_s": 23.1, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 5313, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00208", "task": "L3_046", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "show_all", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L3_046|claude-opus-5-5|1", "design": "f1", "success": true, "score": 1.4327, "status": "ok", "correct": true, "speedup": 1.4327, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 144.0, "calls": 2, "tokens_in": 17418, "tokens_out": 7398, "ctx_chars": 10384, "ctx_sha": "4573cb1c62d0cdf8", "late_delivered": null, "t_end": 1791553730.0128846, "backend": "triton", "own_frac": 0.0803, "ref_ms": 4.07871, "cand_ms": 2.84694, "endpoint": "endpoint-26", "r_wall_s": 67.35, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 5226, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00199", "task": "L2_071", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "late_show_all", "delivery": "late", "rep": 0, "p_arm": 0.111111, "block": "L2_071|claude-opus-5-5|0", "design": "f1", "success": true, "score": 2.1882, "status": "ok", "correct": true, "speedup": 2.1882, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 235.9, "calls": 3, "tokens_in": 23207, "tokens_out": 10575, "ctx_chars": 11132, "ctx_sha": "4caa9b6e3a0949e1", "late_delivered": true, "t_end": 1791553730.6045885, "backend": "cuda", "own_frac": 1.0, "ref_ms": 3.66151, "cand_ms": 1.67333, "endpoint": "endpoint-11", "r_wall_s": 71.54, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 5357, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00200", "task": "L2_027", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_approach", "label": "no_approach", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_027|claude-opus-5-5|0", "design": "f1", "success": true, "score": 3.1203, "status": "ok", "correct": true, "speedup": 3.1203, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 222.1, "calls": 2, "tokens_in": 9617, "tokens_out": 13628, "ctx_chars": 3080, "ctx_sha": "7364d33fba159fed", "late_delivered": null, "t_end": 1791553733.0466068, "backend": "cuda", "own_frac": 0.0006, "ref_ms": 20.09568, "cand_ms": 6.44039, "endpoint": "endpoint-09", "r_wall_s": 49.05, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 10396, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00201", "task": "L1_077", "split": "calib", "reader": "claude-opus-5-5", "arm": "none", "label": "none", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L1_077|claude-opus-5-5|1", "design": "f1", "success": true, "score": 2.0491, "status": "ok", "correct": true, "speedup": 2.0491, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 218.5, "calls": 2, "tokens_in": 6656, "tokens_out": 6488, "ctx_chars": 0, "ctx_sha": "e3b0c44298fc1c14", "late_delivered": null, "t_end": 1791553736.3206685, "backend": "cuda", "own_frac": 0.2114, "ref_ms": 8.63081, "cand_ms": 4.21194, "endpoint": "endpoint-17", "r_wall_s": 87.63, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 4449, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00218", "task": "L2_008", "split": "calib", "reader": "claude-opus-5-5", "arm": "reader_relative", "label": "reader_relative", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_008|claude-opus-5-5|1", "design": "f1", "success": true, "score": 1.072, "status": "ok", "correct": true, "speedup": 1.072, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 69.0, "calls": 2, "tokens_in": 15504, "tokens_out": 3708, "ctx_chars": 11693, "ctx_sha": "e1775af2346a3b0a", "late_delivered": null, "t_end": 1791553736.9050188, "backend": "triton", "own_frac": 0.0227, "ref_ms": 12.67736, "cand_ms": 11.82624, "endpoint": "endpoint-14", "r_wall_s": 20.82, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 2512, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00198", "task": "L1_065", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "late_show_all", "delivery": "late", "rep": 0, "p_arm": 0.111111, "block": "L1_065|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.8109, "status": "ok", "correct": true, "speedup": 1.8109, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 250.6, "calls": 2, "tokens_in": 12746, "tokens_out": 12475, "ctx_chars": 11175, "ctx_sha": "91e4a03c424736b2", "late_delivered": true, "t_end": 1791553744.566984, "backend": "cuda", "own_frac": 0.9998, "ref_ms": 44.15902, "cand_ms": 24.38497, "endpoint": "endpoint-26", "r_wall_s": 82.89, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 8582, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00221", "task": "L2_035", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_numbers", "label": "no_numbers", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_035|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.269, "status": "ok", "correct": true, "speedup": 1.269, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 71.9, "calls": 2, "tokens_in": 13141, "tokens_out": 2196, "ctx_chars": 9830, "ctx_sha": "60e5a3b108cbd2cf", "late_delivered": null, "t_end": 1791553754.6632886, "backend": "triton", "own_frac": 0.0607, "ref_ms": 17.55962, "cand_ms": 13.83683, "endpoint": "endpoint-16", "r_wall_s": 36.27, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 1263, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00210", "task": "L2_071", "split": "calib", "reader": "claude-opus-5-5", "arm": "reader_relative", "label": "reader_relative", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_071|claude-opus-5-5|1", "design": "f1", "success": true, "score": 1.2521, "status": "ok", "correct": true, "speedup": 1.2521, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 148.5, "calls": 2, "tokens_in": 15062, "tokens_out": 3689, "ctx_chars": 11758, "ctx_sha": "d74bf4f221ceed35", "late_delivered": null, "t_end": 1791553760.1230202, "backend": "triton", "own_frac": 1.0, "ref_ms": 3.64136, "cand_ms": 2.9081, "endpoint": "endpoint-24", "r_wall_s": 63.25, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 2547, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00180", "task": "L1_080", "split": "calib", "reader": "claude-opus-5-5", "arm": "frontier_gated", "label": "frontier_gated", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L1_080|claude-opus-5-5|0", "design": "f1", "success": false, "score": 0.9045, "status": "ok", "correct": true, "speedup": 0.9045, "infra": false, "infra_reruns": 0, "error": "ok", "wall_s": 397.3, "calls": 2, "tokens_in": 14290, "tokens_out": 10042, "ctx_chars": 7816, "ctx_sha": "015bd136869dcc12", "late_delivered": null, "t_end": 1791553768.8855746, "backend": "cuda", "own_frac": 0.9997, "ref_ms": 20.10384, "cand_ms": 22.22654, "endpoint": "endpoint-19", "r_wall_s": 120.65, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 7002, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00215", "task": "L2_024", "split": "calib", "reader": "claude-opus-5-5", "arm": "none", "label": "none", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_024|claude-opus-5-5|1", "design": "f1", "success": false, "score": 0.4586, "status": "ok", "correct": true, "speedup": 0.4586, "infra": false, "infra_reruns": 0, "error": "ok", "wall_s": 123.5, "calls": 3, "tokens_in": 12916, "tokens_out": 8375, "ctx_chars": 0, "ctx_sha": "e3b0c44298fc1c14", "late_delivered": null, "t_end": 1791553782.7478786, "backend": "cuda", "own_frac": 0.999, "ref_ms": 2.38555, "cand_ms": 5.20125, "endpoint": "endpoint-07", "r_wall_s": 22.17, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 4109, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00216", "task": "L2_046", "split": "calib", "reader": "claude-opus-5-5", "arm": "lessons_only", "label": "lessons_only", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_046|claude-opus-5-5|1", "design": "f1", "success": true, "score": 1.4036, "status": "ok", "correct": true, "speedup": 1.4036, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 118.3, "calls": 2, "tokens_in": 6896, "tokens_out": 4453, "ctx_chars": 1982, "ctx_sha": "54e16caf9bef9eec", "late_delivered": null, "t_end": 1791553783.7516494, "backend": "cuda", "own_frac": 0.0668, "ref_ms": 19.11258, "cand_ms": 13.61729, "endpoint": "endpoint-01", "r_wall_s": 54.33, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 2730, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00187", "task": "L3_039", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_numbers", "label": "no_numbers", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L3_039|claude-opus-5-5|0", "design": "f1", "success": true, "score": 2.9855, "status": "ok", "correct": true, "speedup": 2.9855, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 373.8, "calls": 3, "tokens_in": 30785, "tokens_out": 25343, "ctx_chars": 8204, "ctx_sha": "83777d4de17c3440", "late_delivered": null, "t_end": 1791553790.8972669, "backend": "cuda", "own_frac": 0.9901, "ref_ms": 31.90452, "cand_ms": 10.68635, "endpoint": "endpoint-18", "r_wall_s": 51.02, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 16423, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00214", "task": "L2_085", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "late_show_all", "delivery": "late", "rep": 1, "p_arm": 0.111111, "block": "L2_085|claude-opus-5-5|1", "design": "f1", "success": true, "score": 2.8951, "status": "ok", "correct": true, "speedup": 2.8951, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 153.8, "calls": 2, "tokens_in": 13685, "tokens_out": 12941, "ctx_chars": 11886, "ctx_sha": "8cbbd063f294068f", "late_delivered": true, "t_end": 1791553797.2878063, "backend": "cuda", "own_frac": 0.055, "ref_ms": 4.73417, "cand_ms": 1.63521, "endpoint": "endpoint-19", "r_wall_s": 21.08, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 8992, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00206", "task": "L3_038", "split": "calib", "reader": "claude-opus-5-5", "arm": "reader_relative", "label": "reader_relative", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L3_038|claude-opus-5-5|0", "design": "f1", "success": true, "score": 3.9965, "status": "ok", "correct": true, "speedup": 3.9965, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 228.0, "calls": 3, "tokens_in": 28308, "tokens_out": 19462, "ctx_chars": 8756, "ctx_sha": "26dc0616aa00d951", "late_delivered": null, "t_end": 1791553798.5166519, "backend": "cuda", "own_frac": 0.8339, "ref_ms": 78.77552, "cand_ms": 19.71137, "endpoint": "endpoint-11", "r_wall_s": 19.27, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 12608, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00211", "task": "L3_008", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_numbers", "label": "no_numbers", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L3_008|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.0933, "status": "ok", "correct": true, "speedup": 1.0933, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 211.5, "calls": 2, "tokens_in": 14510, "tokens_out": 18752, "ctx_chars": 6490, "ctx_sha": "00c26ca778818693", "late_delivered": null, "t_end": 1791553824.9426794, "backend": "triton", "own_frac": 0.166, "ref_ms": 5.24213, "cand_ms": 4.79474, "endpoint": "endpoint-05", "r_wall_s": 31.93, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 14245, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00220", "task": "L2_052", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_approach", "label": "no_approach", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_052|claude-opus-5-5|1", "design": "f1", "success": true, "score": 1.4003, "status": "ok", "correct": true, "speedup": 1.4003, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 150.0, "calls": 2, "tokens_in": 8387, "tokens_out": 6986, "ctx_chars": 3080, "ctx_sha": "7364d33fba159fed", "late_delivered": null, "t_end": 1791553830.1342597, "backend": "triton", "own_frac": 0.1539, "ref_ms": 11.00422, "cand_ms": 7.85844, "endpoint": "endpoint-24", "r_wall_s": 56.18, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 4816, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00229", "task": "L2_067", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_approach", "label": "no_approach", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_067|claude-opus-5-5|1", "design": "f1", "success": true, "score": 2.3095, "status": "ok", "correct": true, "speedup": 2.3095, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 112.0, "calls": 2, "tokens_in": 8575, "tokens_out": 9020, "ctx_chars": 3080, "ctx_sha": "7364d33fba159fed", "late_delivered": null, "t_end": 1791553856.6089966, "backend": "cuda", "own_frac": 0.9997, "ref_ms": 12.83502, "cand_ms": 5.55747, "endpoint": "endpoint-11", "r_wall_s": 19.08, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 6644, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00212", "task": "L2_082", "split": "calib", "reader": "claude-opus-5-5", "arm": "lessons_only", "label": "lessons_only", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_082|claude-opus-5-5|0", "design": "f1", "success": true, "score": 4.4046, "status": "ok", "correct": true, "speedup": 4.4046, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 250.3, "calls": 3, "tokens_in": 18246, "tokens_out": 15865, "ctx_chars": 1982, "ctx_sha": "54e16caf9bef9eec", "late_delivered": null, "t_end": 1791553865.479409, "backend": "cuda", "own_frac": 1.0, "ref_ms": 19.49564, "cand_ms": 4.42624, "endpoint": "endpoint-24", "r_wall_s": 30.42, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 9343, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00230", "task": "L2_035", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "late_show_all", "delivery": "late", "rep": 0, "p_arm": 0.111111, "block": "L2_035|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.2761, "status": "ok", "correct": true, "speedup": 1.2761, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 112.5, "calls": 2, "tokens_in": 9477, "tokens_out": 4038, "ctx_chars": 9835, "ctx_sha": "6aedb10d18ef2e36", "late_delivered": true, "t_end": 1791553867.1647549, "backend": "cuda", "own_frac": 0.0731, "ref_ms": 17.3707, "cand_ms": 13.61227, "endpoint": "endpoint-26", "r_wall_s": 49.26, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 2651, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00217", "task": "L2_085", "split": "calib", "reader": "claude-opus-5-5", "arm": "none", "label": "none", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_085|claude-opus-5-5|0", "design": "f1", "success": true, "score": 2.992, "status": "ok", "correct": true, "speedup": 2.992, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 201.2, "calls": 2, "tokens_in": 8441, "tokens_out": 16180, "ctx_chars": 0, "ctx_sha": "e3b0c44298fc1c14", "late_delivered": null, "t_end": 1791553868.7140377, "backend": "cuda", "own_frac": 0.9985, "ref_ms": 4.76032, "cand_ms": 1.59104, "endpoint": "endpoint-17", "r_wall_s": 25.18, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 12088, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00236", "task": "L2_008", "split": "calib", "reader": "claude-opus-5-5", "arm": "reader_relative", "label": "reader_relative", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_008|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.0727, "status": "ok", "correct": true, "speedup": 1.0727, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 73.6, "calls": 2, "tokens_in": 15537, "tokens_out": 3809, "ctx_chars": 11693, "ctx_sha": "e1775af2346a3b0a", "late_delivered": null, "t_end": 1791553870.8912318, "backend": "triton", "own_frac": 0.023, "ref_ms": 12.70078, "cand_ms": 11.83998, "endpoint": "endpoint-01", "r_wall_s": 13.02, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 2615, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00219", "task": "L1_078", "split": "calib", "reader": "claude-opus-5-5", "arm": "lessons_only", "label": "lessons_only", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L1_078|claude-opus-5-5|1", "design": "f1", "success": true, "score": 3.1234, "status": "ok", "correct": true, "speedup": 3.1234, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 198.2, "calls": 2, "tokens_in": 8493, "tokens_out": 9504, "ctx_chars": 1982, "ctx_sha": "54e16caf9bef9eec", "late_delivered": null, "t_end": 1791553877.3184402, "backend": "cuda", "own_frac": 0.9992, "ref_ms": 35.48968, "cand_ms": 11.36267, "endpoint": "endpoint-18", "r_wall_s": 82.26, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 6997, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00227", "task": "L1_083", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_numbers", "label": "no_numbers", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L1_083|claude-opus-5-5|1", "design": "f1", "success": true, "score": 2.5769, "status": "ok", "correct": true, "speedup": 2.5769, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 143.9, "calls": 3, "tokens_in": 24581, "tokens_out": 5211, "ctx_chars": 10351, "ctx_sha": "31905f761c946397", "late_delivered": null, "t_end": 1791553880.1741335, "backend": "triton", "own_frac": 1.0, "ref_ms": 2.5898, "cand_ms": 1.00501, "endpoint": "endpoint-14", "r_wall_s": 53.48, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "triton", "tok_out_first": 2289, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "runtime_error"} {"ep": "f1-00232", "task": "L2_071", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_numbers", "label": "no_numbers", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_071|claude-opus-5-5|1", "design": "f1", "success": true, "score": 1.2521, "status": "ok", "correct": true, "speedup": 1.2521, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 120.1, "calls": 2, "tokens_in": 13669, "tokens_out": 2055, "ctx_chars": 11127, "ctx_sha": "e2f40f0540528624", "late_delivered": null, "t_end": 1791553889.03509, "backend": "triton", "own_frac": 0.2369, "ref_ms": 3.64333, "cand_ms": 2.90984, "endpoint": "endpoint-26", "r_wall_s": 71.59, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 1448, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00207", "task": "L3_038", "split": "calib", "reader": "claude-opus-5-5", "arm": "frontier_gated", "label": "frontier_gated", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L3_038|claude-opus-5-5|1", "design": "f1", "success": true, "score": 4.0483, "status": "ok", "correct": true, "speedup": 4.0483, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 313.6, "calls": 3, "tokens_in": 19057, "tokens_out": 15895, "ctx_chars": 2260, "ctx_sha": "9b6af621f8f8a88b", "late_delivered": null, "t_end": 1791553897.078643, "backend": "cuda", "own_frac": 0.8309, "ref_ms": 78.70648, "cand_ms": 19.44171, "endpoint": "endpoint-10", "r_wall_s": 46.49, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 9430, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00167", "task": "L3_024", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_approach", "label": "no_approach", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L3_024|claude-opus-5-5|1", "design": "f1", "success": false, "score": 0.8744, "status": "ok", "correct": true, "speedup": 0.8744, "infra": false, "infra_reruns": 0, "error": "ok", "wall_s": 616.7, "calls": 4, "tokens_in": 63562, "tokens_out": 59238, "ctx_chars": 2932, "ctx_sha": "e5406f0f931ba9ce", "late_delivered": null, "t_end": 1791553898.7641108, "backend": "cuda", "own_frac": 0.0287, "ref_ms": 3.59646, "cand_ms": 4.11297, "endpoint": "endpoint-11", "r_wall_s": 35.66, "flags": [], "submit_turn": 3, "turns": 4, "tech_src": "cuda", "tok_out_first": 21422, "stops": ["end_turn", "end_turn", "end_turn", "end_turn"], "first_check": "runtime_error"} {"ep": "f1-00223", "task": "L2_024", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_approach", "label": "no_approach", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_024|claude-opus-5-5|0", "design": "f1", "success": false, "score": 0.2607, "status": "ok", "correct": true, "speedup": 0.2607, "infra": false, "infra_reruns": 0, "error": "ok", "wall_s": 173.9, "calls": 3, "tokens_in": 15700, "tokens_out": 8791, "ctx_chars": 3080, "ctx_sha": "7364d33fba159fed", "late_delivered": null, "t_end": 1791553901.7977772, "backend": "cuda", "own_frac": 0.9995, "ref_ms": 2.42544, "cand_ms": 9.3025, "endpoint": "endpoint-11", "r_wall_s": 39.14, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 4853, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00231", "task": "L2_018", "split": "calib", "reader": "claude-opus-5-5", "arm": "frontier_gated", "label": "frontier_gated", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_018|claude-opus-5-5|1", "design": "f1", "success": true, "score": 7.2299, "status": "ok", "correct": true, "speedup": 7.2299, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 143.6, "calls": 3, "tokens_in": 11741, "tokens_out": 8149, "ctx_chars": 2990, "ctx_sha": "2d414f7ee5025848", "late_delivered": null, "t_end": 1791553903.7371602, "backend": "triton", "own_frac": 0.0273, "ref_ms": 8.02938, "cand_ms": 1.11058, "endpoint": "endpoint-17", "r_wall_s": 12.68, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "triton", "tok_out_first": 1807, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "incorrect"} {"ep": "f1-00226", "task": "L2_052", "split": "calib", "reader": "claude-opus-5-5", "arm": "frontier_gated", "label": "frontier_gated", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_052|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.3981, "status": "ok", "correct": true, "speedup": 1.3981, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 173.0, "calls": 2, "tokens_in": 8422, "tokens_out": 5236, "ctx_chars": 3639, "ctx_sha": "fc6e42d43dfb4e75", "late_delivered": null, "t_end": 1791553906.076794, "backend": "triton", "own_frac": 0.1436, "ref_ms": 11.02984, "cand_ms": 7.88893, "endpoint": "endpoint-19", "r_wall_s": 85.28, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 3344, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00225", "task": "L1_077", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_numbers", "label": "no_numbers", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L1_077|claude-opus-5-5|0", "design": "f1", "success": false, "score": 0.6304, "status": "ok", "correct": true, "speedup": 0.6304, "infra": false, "infra_reruns": 0, "error": "ok", "wall_s": 176.3, "calls": 2, "tokens_in": 15667, "tokens_out": 5425, "ctx_chars": 11170, "ctx_sha": "a5aff0bf6f7d7e8d", "late_delivered": null, "t_end": 1791553906.8741271, "backend": "triton", "own_frac": 0.7607, "ref_ms": 8.62475, "cand_ms": 13.68188, "endpoint": "endpoint-23", "r_wall_s": 61.53, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 3965, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00222", "task": "L1_077", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_numbers", "label": "no_numbers", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L1_077|claude-opus-5-5|1", "design": "f1", "success": true, "score": 2.1279, "status": "ok", "correct": true, "speedup": 2.1279, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 222.3, "calls": 2, "tokens_in": 15915, "tokens_out": 6585, "ctx_chars": 11170, "ctx_sha": "a5aff0bf6f7d7e8d", "late_delivered": null, "t_end": 1791553922.6238155, "backend": "triton", "own_frac": 0.1308, "ref_ms": 8.72041, "cand_ms": 4.09822, "endpoint": "endpoint-16", "r_wall_s": 120.77, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 4869, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00233", "task": "L2_057", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_approach", "label": "no_approach", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_057|claude-opus-5-5|0", "design": "f1", "success": true, "score": 3.3696, "status": "ok", "correct": true, "speedup": 3.3696, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 153.5, "calls": 2, "tokens_in": 8077, "tokens_out": 6748, "ctx_chars": 3080, "ctx_sha": "7364d33fba159fed", "late_delivered": null, "t_end": 1791553936.281674, "backend": "cuda", "own_frac": 1.0, "ref_ms": 6.3129, "cand_ms": 1.87348, "endpoint": "endpoint-09", "r_wall_s": 68.35, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 4603, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00238", "task": "L1_050", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_approach", "label": "no_approach", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L1_050|claude-opus-5-5|1", "design": "f1", "success": true, "score": 1.5541, "status": "ok", "correct": true, "speedup": 1.5541, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 115.7, "calls": 2, "tokens_in": 7469, "tokens_out": 5601, "ctx_chars": 3080, "ctx_sha": "7364d33fba159fed", "late_delivered": null, "t_end": 1791553940.6542966, "backend": "triton", "own_frac": 0.9982, "ref_ms": 5.31756, "cand_ms": 3.4216, "endpoint": "endpoint-18", "r_wall_s": 42.87, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 4148, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00239", "task": "L1_050", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "show_all", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L1_050|claude-opus-5-5|1", "design": "f1", "success": true, "score": 1.297, "status": "ok", "correct": true, "speedup": 1.297, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 121.5, "calls": 2, "tokens_in": 16853, "tokens_out": 6582, "ctx_chars": 13278, "ctx_sha": "cb78f644acd17887", "late_delivered": null, "t_end": 1791553951.5875778, "backend": "triton", "own_frac": 0.9883, "ref_ms": 5.33584, "cand_ms": 4.11397, "endpoint": "endpoint-07", "r_wall_s": 45.14, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 4602, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00228", "task": "L1_062", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_numbers", "label": "no_numbers", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L1_062|claude-opus-5-5|1", "design": "f1", "success": false, "score": 0.3065, "status": "ok", "correct": true, "speedup": 0.3065, "infra": false, "infra_reruns": 0, "error": "ok", "wall_s": 224.7, "calls": 2, "tokens_in": 16187, "tokens_out": 7674, "ctx_chars": 10859, "ctx_sha": "07dfed07acc9518f", "late_delivered": null, "t_end": 1791553961.557269, "backend": "triton", "own_frac": 0.9999, "ref_ms": 10.02863, "cand_ms": 32.72328, "endpoint": "endpoint-25", "r_wall_s": 102.03, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 5327, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00245", "task": "L2_011", "split": "calib", "reader": "claude-opus-5-5", "arm": "frontier_gated", "label": "frontier_gated", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_011|claude-opus-5-5|1", "design": "f1", "success": true, "score": 1.0482, "status": "ok", "correct": true, "speedup": 1.0482, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 84.9, "calls": 2, "tokens_in": 11054, "tokens_out": 6003, "ctx_chars": 5772, "ctx_sha": "f29083df0679b045", "late_delivered": null, "t_end": 1791553962.2573905, "backend": "triton", "own_frac": 0.0289, "ref_ms": 27.93073, "cand_ms": 26.64662, "endpoint": "endpoint-23", "r_wall_s": 21.2, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 4135, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00153", "task": "L3_024", "split": "calib", "reader": "claude-opus-5-5", "arm": "none", "label": "none", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L3_024|claude-opus-5-5|0", "design": "f1", "success": false, "score": 0.8635, "status": "ok", "correct": true, "speedup": 0.8635, "infra": false, "infra_reruns": 0, "error": "ok", "wall_s": 787.8, "calls": 5, "tokens_in": 98441, "tokens_out": 76254, "ctx_chars": 0, "ctx_sha": "e3b0c44298fc1c14", "late_delivered": null, "t_end": 1791553978.2883635, "backend": "cuda", "own_frac": 0.0218, "ref_ms": 3.45914, "cand_ms": 4.00572, "endpoint": "endpoint-05", "r_wall_s": 20.56, "flags": [], "submit_turn": 4, "turns": 5, "tech_src": "cuda", "tok_out_first": 31394, "stops": ["end_turn", "end_turn", "end_turn", "end_turn", "end_turn"], "first_check": "runtime_error"} {"ep": "f1-00235", "task": "L1_050", "split": "calib", "reader": "claude-opus-5-5", "arm": "none", "label": "none", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L1_050|claude-opus-5-5|1", "design": "f1", "success": true, "score": 1.3075, "status": "ok", "correct": true, "speedup": 1.3075, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 188.7, "calls": 2, "tokens_in": 5881, "tokens_out": 10233, "ctx_chars": 0, "ctx_sha": "e3b0c44298fc1c14", "late_delivered": null, "t_end": 1791553979.596032, "backend": "cuda", "own_frac": 0.9985, "ref_ms": 5.31728, "cand_ms": 4.06678, "endpoint": "endpoint-16", "r_wall_s": 55.96, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 7775, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00237", "task": "L3_042", "split": "calib", "reader": "claude-opus-5-5", "arm": "frontier_gated", "label": "frontier_gated", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L3_042|claude-opus-5-5|1", "design": "f1", "success": true, "score": 3.6513, "status": "ok", "correct": true, "speedup": 3.6513, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 188.8, "calls": 3, "tokens_in": 18702, "tokens_out": 14465, "ctx_chars": 2260, "ctx_sha": "9b6af621f8f8a88b", "late_delivered": null, "t_end": 1791553987.291251, "backend": "cuda", "own_frac": 0.8497, "ref_ms": 60.98984, "cand_ms": 16.70355, "endpoint": "endpoint-10", "r_wall_s": 27.03, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 8159, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00234", "task": "L3_046", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "late_show_all", "delivery": "late", "rep": 1, "p_arm": 0.111111, "block": "L3_046|claude-opus-5-5|1", "design": "f1", "success": true, "score": 2.1037, "status": "ok", "correct": true, "speedup": 2.1037, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 212.2, "calls": 3, "tokens_in": 29577, "tokens_out": 15169, "ctx_chars": 10384, "ctx_sha": "4573cb1c62d0cdf8", "late_delivered": true, "t_end": 1791553995.955795, "backend": "cuda", "own_frac": 0.4866, "ref_ms": 4.06158, "cand_ms": 1.93072, "endpoint": "endpoint-17", "r_wall_s": 47.56, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 7741, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00253", "task": "L2_046", "split": "calib", "reader": "claude-opus-5-5", "arm": "fg_target", "label": "fg_target", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_046|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.3978, "status": "ok", "correct": true, "speedup": 1.3978, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 90.0, "calls": 2, "tokens_in": 7457, "tokens_out": 2668, "ctx_chars": 3639, "ctx_sha": "fc6e42d43dfb4e75", "late_delivered": null, "t_end": 1791553996.8672702, "backend": "triton", "own_frac": 0.0678, "ref_ms": 19.32173, "cand_ms": 13.82277, "endpoint": "endpoint-10", "r_wall_s": 35.7, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 1769, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00250", "task": "L2_035", "split": "calib", "reader": "claude-opus-5-5", "arm": "frontier_gated", "label": "frontier_gated", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_035|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.2725, "status": "ok", "correct": true, "speedup": 1.2725, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 96.9, "calls": 2, "tokens_in": 9550, "tokens_out": 2967, "ctx_chars": 5671, "ctx_sha": "7f0e3598c3810513", "late_delivered": null, "t_end": 1791553998.7044592, "backend": "triton", "own_frac": 0.0337, "ref_ms": 17.41046, "cand_ms": 13.68229, "endpoint": "endpoint-17", "r_wall_s": 52.21, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 2011, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00254", "task": "L2_025", "split": "calib", "reader": "claude-opus-5-5", "arm": "reader_relative", "label": "reader_relative", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_025|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.3728, "status": "ok", "correct": true, "speedup": 1.3728, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 98.8, "calls": 2, "tokens_in": 13982, "tokens_out": 4942, "ctx_chars": 10713, "ctx_sha": "6db24bfdf4423291", "late_delivered": null, "t_end": 1791554021.385576, "backend": "triton", "own_frac": 1.0, "ref_ms": 19.23877, "cand_ms": 14.01452, "endpoint": "endpoint-14", "r_wall_s": 22.97, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 3692, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00257", "task": "L2_046", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_approach", "label": "no_approach", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_046|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.3898, "status": "ok", "correct": true, "speedup": 1.3898, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 81.3, "calls": 2, "tokens_in": 7370, "tokens_out": 3245, "ctx_chars": 3080, "ctx_sha": "7364d33fba159fed", "late_delivered": null, "t_end": 1791554032.8518474, "backend": "triton", "own_frac": 0.0668, "ref_ms": 19.33392, "cand_ms": 13.91136, "endpoint": "endpoint-10", "r_wall_s": 33.54, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 2166, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00252", "task": "L1_077", "split": "calib", "reader": "claude-opus-5-5", "arm": "reader_relative", "label": "reader_relative", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L1_077|claude-opus-5-5|1", "design": "f1", "success": true, "score": 1.9895, "status": "ok", "correct": true, "speedup": 1.9895, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 132.9, "calls": 2, "tokens_in": 16424, "tokens_out": 5648, "ctx_chars": 11807, "ctx_sha": "a79e3b8677344448", "late_delivered": null, "t_end": 1791554038.969173, "backend": "triton", "own_frac": 0.2188, "ref_ms": 8.60353, "cand_ms": 4.3244, "endpoint": "endpoint-01", "r_wall_s": 60.7, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 4056, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00246", "task": "L1_065", "split": "calib", "reader": "claude-opus-5-5", "arm": "none", "label": "none", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L1_065|claude-opus-5-5|1", "design": "f1", "success": true, "score": 2.0705, "status": "ok", "correct": true, "speedup": 2.0705, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 160.0, "calls": 2, "tokens_in": 6337, "tokens_out": 7385, "ctx_chars": 0, "ctx_sha": "e3b0c44298fc1c14", "late_delivered": null, "t_end": 1791554040.2070627, "backend": "triton", "own_frac": 0.9998, "ref_ms": 44.2128, "cand_ms": 21.35338, "endpoint": "endpoint-23", "r_wall_s": 68.22, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 5790, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00242", "task": "L2_008", "split": "calib", "reader": "claude-opus-5-5", "arm": "fg_target", "label": "fg_target", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_008|claude-opus-5-5|1", "design": "f1", "success": true, "score": 4.5742, "status": "ok", "correct": true, "speedup": 4.5742, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 173.5, "calls": 3, "tokens_in": 19091, "tokens_out": 13725, "ctx_chars": 3639, "ctx_sha": "fc6e42d43dfb4e75", "late_delivered": null, "t_end": 1791554040.6583872, "backend": "cuda", "own_frac": 0.9958, "ref_ms": 12.67155, "cand_ms": 2.77024, "endpoint": "endpoint-19", "r_wall_s": 28.17, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 8230, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00241", "task": "L3_004", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "late_show_all", "delivery": "late", "rep": 0, "p_arm": 0.111111, "block": "L3_004|claude-opus-5-5|0", "design": "f1", "success": true, "score": 3.8574, "status": "ok", "correct": true, "speedup": 3.8574, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 190.4, "calls": 3, "tokens_in": 26186, "tokens_out": 16533, "ctx_chars": 6513, "ctx_sha": "1aa698d3a9589f25", "late_delivered": true, "t_end": 1791554055.8680973, "backend": "cuda", "own_frac": 1.0, "ref_ms": 2.3073, "cand_ms": 0.59815, "endpoint": "endpoint-18", "r_wall_s": 20.23, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 7199, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00249", "task": "L3_038", "split": "calib", "reader": "claude-opus-5-5", "arm": "reader_relative", "label": "reader_relative", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L3_038|claude-opus-5-5|1", "design": "f1", "success": false, "score": 0.8833, "status": "ok", "correct": true, "speedup": 0.8833, "infra": false, "infra_reruns": 0, "error": "ok", "wall_s": 161.0, "calls": 2, "tokens_in": 14489, "tokens_out": 8524, "ctx_chars": 8756, "ctx_sha": "26dc0616aa00d951", "late_delivered": null, "t_end": 1791554059.7888587, "backend": "triton", "own_frac": 0.0005, "ref_ms": 116.89573, "cand_ms": 132.34574, "endpoint": "endpoint-11", "r_wall_s": 37.93, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 5887, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00240", "task": "L2_027", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "show_all", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_027|claude-opus-5-5|1", "design": "f1", "success": true, "score": 3.5202, "status": "ok", "correct": true, "speedup": 3.5202, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 212.4, "calls": 2, "tokens_in": 16559, "tokens_out": 8892, "ctx_chars": 11535, "ctx_sha": "114ec077101d70e2", "late_delivered": null, "t_end": 1791554068.9748428, "backend": "triton", "own_frac": 1.0, "ref_ms": 20.01403, "cand_ms": 5.68545, "endpoint": "endpoint-05", "r_wall_s": 61.87, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 6356, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00251", "task": "L2_025", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_approach", "label": "no_approach", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_025|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.9962, "status": "ok", "correct": true, "speedup": 1.9962, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 166.0, "calls": 2, "tokens_in": 8492, "tokens_out": 9260, "ctx_chars": 3080, "ctx_sha": "7364d33fba159fed", "late_delivered": null, "t_end": 1791554069.7428265, "backend": "cuda", "own_frac": 0.9997, "ref_ms": 19.36408, "cand_ms": 9.70043, "endpoint": "endpoint-24", "r_wall_s": 60.38, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 6909, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00224", "task": "L3_039", "split": "calib", "reader": "claude-opus-5-5", "arm": "lessons_only", "label": "lessons_only", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L3_039|claude-opus-5-5|1", "design": "f1", "success": true, "score": 8.3869, "status": "ok", "correct": true, "speedup": 8.3869, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 354.1, "calls": 3, "tokens_in": 20107, "tokens_out": 18624, "ctx_chars": 1982, "ctx_sha": "54e16caf9bef9eec", "late_delivered": null, "t_end": 1791554084.0917215, "backend": "cuda", "own_frac": 0.9995, "ref_ms": 32.08403, "cand_ms": 3.82548, "endpoint": "endpoint-24", "r_wall_s": 74.88, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 11170, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00258", "task": "L2_025", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "late_show_all", "delivery": "late", "rep": 0, "p_arm": 0.111111, "block": "L2_025|claude-opus-5-5|0", "design": "f1", "success": true, "score": 2.2397, "status": "ok", "correct": true, "speedup": 2.2397, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 127.2, "calls": 2, "tokens_in": 10288, "tokens_out": 8582, "ctx_chars": 10087, "ctx_sha": "fc8a2930809fd512", "late_delivered": true, "t_end": 1791554088.7129781, "backend": "cuda", "own_frac": 0.9996, "ref_ms": 19.31296, "cand_ms": 8.62309, "endpoint": "endpoint-23", "r_wall_s": 23.27, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 6098, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00248", "task": "L3_042", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_approach", "label": "no_approach", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L3_042|claude-opus-5-5|0", "design": "f1", "success": true, "score": 2.3245, "status": "ok", "correct": true, "speedup": 2.3245, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 194.2, "calls": 3, "tokens_in": 20683, "tokens_out": 16285, "ctx_chars": 3355, "ctx_sha": "cc270772e526c41f", "late_delivered": null, "t_end": 1791554091.2862718, "backend": "cuda", "own_frac": 0.9053, "ref_ms": 61.13075, "cand_ms": 26.29875, "endpoint": "endpoint-18", "r_wall_s": 20.52, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 9400, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00244", "task": "L3_049", "split": "calib", "reader": "claude-opus-5-5", "arm": "frontier_gated", "label": "frontier_gated", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L3_049|claude-opus-5-5|0", "design": "f1", "success": true, "score": 2.4631, "status": "ok", "correct": true, "speedup": 2.4631, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 223.7, "calls": 4, "tokens_in": 30099, "tokens_out": 14556, "ctx_chars": 2578, "ctx_sha": "7b4237bf96a5394d", "late_delivered": null, "t_end": 1791554094.589286, "backend": "cuda", "own_frac": 0.4264, "ref_ms": 11.98027, "cand_ms": 4.86388, "endpoint": "endpoint-16", "r_wall_s": 28.66, "flags": [], "submit_turn": 3, "turns": 4, "tech_src": "cuda", "tok_out_first": 5216, "stops": ["end_turn", "end_turn", "end_turn", "end_turn"], "first_check": "incorrect"} {"ep": "f1-00261", "task": "L2_067", "split": "calib", "reader": "claude-opus-5-5", "arm": "reader_relative", "label": "reader_relative", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_067|claude-opus-5-5|0", "design": "f1", "success": true, "score": 2.2802, "status": "ok", "correct": true, "speedup": 2.2802, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 126.9, "calls": 2, "tokens_in": 16307, "tokens_out": 9769, "ctx_chars": 11649, "ctx_sha": "4bc9a6c4c4f94c6f", "late_delivered": null, "t_end": 1791554106.4905205, "backend": "cuda", "own_frac": 0.9997, "ref_ms": 12.83935, "cand_ms": 5.6307, "endpoint": "endpoint-24", "r_wall_s": 20.29, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 7399, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00247", "task": "L3_036", "split": "calib", "reader": "claude-opus-5-5", "arm": "fg_target", "label": "fg_target", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L3_036|claude-opus-5-5|0", "design": "f1", "success": true, "score": 7.0272, "status": "ok", "correct": true, "speedup": 7.0272, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 224.5, "calls": 3, "tokens_in": 20208, "tokens_out": 20740, "ctx_chars": 2260, "ctx_sha": "9b6af621f8f8a88b", "late_delivered": null, "t_end": 1791554113.5041773, "backend": "cuda", "own_frac": 0.9611, "ref_ms": 32.5757, "cand_ms": 4.63565, "endpoint": "endpoint-14", "r_wall_s": 18.94, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 13310, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00260", "task": "L2_067", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "late_show_all", "delivery": "late", "rep": 1, "p_arm": 0.111111, "block": "L2_067|claude-opus-5-5|1", "design": "f1", "success": true, "score": 2.4961, "status": "ok", "correct": true, "speedup": 2.4961, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 170.3, "calls": 3, "tokens_in": 23316, "tokens_out": 12076, "ctx_chars": 11023, "ctx_sha": "04625c0d6e6734eb", "late_delivered": true, "t_end": 1791554148.6376953, "backend": "cuda", "own_frac": 0.9996, "ref_ms": 12.87987, "cand_ms": 5.16006, "endpoint": "endpoint-17", "r_wall_s": 21.38, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 6805, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00266", "task": "L2_011", "split": "calib", "reader": "claude-opus-5-5", "arm": "lessons_only", "label": "lessons_only", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_011|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.0605, "status": "ok", "correct": true, "speedup": 1.0605, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 128.4, "calls": 2, "tokens_in": 8921, "tokens_out": 8661, "ctx_chars": 1982, "ctx_sha": "54e16caf9bef9eec", "late_delivered": null, "t_end": 1791554149.762635, "backend": "cuda", "own_frac": 0.0243, "ref_ms": 27.91125, "cand_ms": 26.32018, "endpoint": "endpoint-07", "r_wall_s": 40.7, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 5444, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00262", "task": "L1_072", "split": "calib", "reader": "claude-opus-5-5", "arm": "none", "label": "none", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L1_072|claude-opus-5-5|1", "design": "f1", "success": true, "score": 3.4852, "status": "ok", "correct": true, "speedup": 3.4852, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 184.0, "calls": 2, "tokens_in": 8476, "tokens_out": 12430, "ctx_chars": 0, "ctx_sha": "e3b0c44298fc1c14", "late_delivered": null, "t_end": 1791554171.276379, "backend": "cuda", "own_frac": 1.0, "ref_ms": 4.37913, "cand_ms": 1.25648, "endpoint": "endpoint-24", "r_wall_s": 33.66, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 8821, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00273", "task": "L3_038", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "show_all", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L3_038|claude-opus-5-5|0", "design": "f1", "success": false, "score": 0.8074, "status": "ok", "correct": true, "speedup": 0.8074, "infra": false, "infra_reruns": 0, "error": "ok", "wall_s": 104.2, "calls": 2, "tokens_in": 13456, "tokens_out": 8393, "ctx_chars": 8224, "ctx_sha": "71fd1a97cbd02e3b", "late_delivered": null, "t_end": 1791554173.1937509, "backend": "triton", "own_frac": 0.0001, "ref_ms": 78.27335, "cand_ms": 96.94945, "endpoint": "endpoint-05", "r_wall_s": 15.57, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 6483, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00271", "task": "L3_046", "split": "calib", "reader": "claude-opus-5-5", "arm": "reader_relative", "label": "reader_relative", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L3_046|claude-opus-5-5|1", "design": "f1", "success": true, "score": 2.2467, "status": "ok", "correct": true, "speedup": 2.2467, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 118.4, "calls": 2, "tokens_in": 17899, "tokens_out": 8218, "ctx_chars": 11007, "ctx_sha": "94e27c56103f4a46", "late_delivered": null, "t_end": 1791554174.2947116, "backend": "triton", "own_frac": 0.4568, "ref_ms": 4.01607, "cand_ms": 1.78754, "endpoint": "endpoint-19", "r_wall_s": 28.84, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 6134, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00263", "task": "L3_042", "split": "calib", "reader": "claude-opus-5-5", "arm": "lessons_only", "label": "lessons_only", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L3_042|claude-opus-5-5|0", "design": "f1", "success": true, "score": 2.705, "status": "ok", "correct": true, "speedup": 2.705, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 182.7, "calls": 3, "tokens_in": 18892, "tokens_out": 14720, "ctx_chars": 1982, "ctx_sha": "54e16caf9bef9eec", "late_delivered": null, "t_end": 1791554178.700118, "backend": "cuda", "own_frac": 0.8845, "ref_ms": 58.0758, "cand_ms": 21.46982, "endpoint": "endpoint-16", "r_wall_s": 21.9, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 8305, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00276", "task": "L2_035", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "late_show_all", "delivery": "late", "rep": 1, "p_arm": 0.111111, "block": "L2_035|claude-opus-5-5|1", "design": "f1", "success": true, "score": 1.2774, "status": "ok", "correct": true, "speedup": 1.2774, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 97.4, "calls": 2, "tokens_in": 9231, "tokens_out": 3542, "ctx_chars": 9835, "ctx_sha": "6aedb10d18ef2e36", "late_delivered": true, "t_end": 1791554186.1347358, "backend": "triton", "own_frac": 0.0577, "ref_ms": 17.40551, "cand_ms": 13.62535, "endpoint": "endpoint-01", "r_wall_s": 36.55, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 2244, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00267", "task": "L2_073", "split": "calib", "reader": "claude-opus-5-5", "arm": "reader_relative", "label": "reader_relative", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_073|claude-opus-5-5|1", "design": "f1", "success": true, "score": 1.6243, "status": "ok", "correct": true, "speedup": 1.6243, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 159.1, "calls": 2, "tokens_in": 13789, "tokens_out": 4535, "ctx_chars": 10198, "ctx_sha": "5a2adf43968d15b9", "late_delivered": null, "t_end": 1791554191.9791577, "backend": "triton", "own_frac": 0.3221, "ref_ms": 5.61583, "cand_ms": 3.45739, "endpoint": "endpoint-23", "r_wall_s": 60.17, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 3128, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00272", "task": "L2_052", "split": "calib", "reader": "claude-opus-5-5", "arm": "none", "label": "none", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_052|claude-opus-5-5|1", "design": "f1", "success": true, "score": 1.4067, "status": "ok", "correct": true, "speedup": 1.4067, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 141.1, "calls": 2, "tokens_in": 5284, "tokens_out": 4657, "ctx_chars": 0, "ctx_sha": "e3b0c44298fc1c14", "late_delivered": null, "t_end": 1791554200.8644547, "backend": "triton", "own_frac": 0.1367, "ref_ms": 11.09718, "cand_ms": 7.88893, "endpoint": "endpoint-18", "r_wall_s": 82.47, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 3138, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00255", "task": "L1_062", "split": "calib", "reader": "claude-opus-5-5", "arm": "reader_relative", "label": "reader_relative", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L1_062|claude-opus-5-5|0", "design": "f1", "success": false, "score": 0.4277, "status": "ok", "correct": true, "speedup": 0.4277, "infra": false, "infra_reruns": 0, "error": "ok", "wall_s": 272.7, "calls": 3, "tokens_in": 32747, "tokens_out": 16947, "ctx_chars": 11490, "ctx_sha": "643e4c44bbeb656a", "late_delivered": null, "t_end": 1791554208.9926407, "backend": "cuda", "own_frac": 0.9998, "ref_ms": 10.07136, "cand_ms": 23.54954, "endpoint": "endpoint-25", "r_wall_s": 75.47, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 9346, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00259", "task": "L3_004", "split": "calib", "reader": "claude-opus-5-5", "arm": "lessons_only", "label": "lessons_only", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L3_004|claude-opus-5-5|0", "design": "f1", "success": true, "score": 6.3348, "status": "ok", "correct": true, "speedup": 6.3348, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 248.0, "calls": 3, "tokens_in": 20341, "tokens_out": 19978, "ctx_chars": 1982, "ctx_sha": "54e16caf9bef9eec", "late_delivered": null, "t_end": 1791554210.213476, "backend": "cuda", "own_frac": 0.6575, "ref_ms": 2.30342, "cand_ms": 0.36362, "endpoint": "endpoint-16", "r_wall_s": 19.51, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 12699, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00275", "task": "L2_018", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "show_all", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_018|claude-opus-5-5|1", "design": "f1", "success": true, "score": 1.0093, "status": "ok", "correct": true, "speedup": 1.0093, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 132.5, "calls": 4, "tokens_in": 33139, "tokens_out": 7639, "ctx_chars": 9898, "ctx_sha": "eb48d1e5953cd4b8", "late_delivered": null, "t_end": 1791554216.590164, "backend": "triton", "own_frac": 0.0029, "ref_ms": 7.9954, "cand_ms": 7.92138, "endpoint": "endpoint-19", "r_wall_s": 13.12, "flags": [], "submit_turn": 3, "turns": 4, "tech_src": "triton", "tok_out_first": 2482, "stops": ["end_turn", "end_turn", "end_turn", "end_turn"], "first_check": "incorrect"} {"ep": "f1-00268", "task": "L1_077", "split": "calib", "reader": "claude-opus-5-5", "arm": "lessons_only", "label": "lessons_only", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L1_077|claude-opus-5-5|1", "design": "f1", "success": true, "score": 1.5696, "status": "ok", "correct": true, "speedup": 1.5696, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 179.4, "calls": 2, "tokens_in": 7550, "tokens_out": 5696, "ctx_chars": 1982, "ctx_sha": "54e16caf9bef9eec", "late_delivered": null, "t_end": 1791554218.3993225, "backend": "triton", "own_frac": 0.2355, "ref_ms": 8.65566, "cand_ms": 5.51459, "endpoint": "endpoint-14", "r_wall_s": 95.77, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 4089, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00270", "task": "L3_039", "split": "calib", "reader": "claude-opus-5-5", "arm": "reader_relative", "label": "reader_relative", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L3_039|claude-opus-5-5|0", "design": "f1", "success": true, "score": 2.4109, "status": "ok", "correct": true, "speedup": 2.4109, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 190.9, "calls": 3, "tokens_in": 27467, "tokens_out": 15596, "ctx_chars": 8756, "ctx_sha": "26dc0616aa00d951", "late_delivered": null, "t_end": 1791554231.5357325, "backend": "cuda", "own_frac": 0.941, "ref_ms": 32.23987, "cand_ms": 13.37254, "endpoint": "endpoint-18", "r_wall_s": 19.72, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 9023, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00281", "task": "L2_008", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "show_all", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_008|claude-opus-5-5|1", "design": "f1", "success": true, "score": 1.0698, "status": "ok", "correct": true, "speedup": 1.0698, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 99.6, "calls": 2, "tokens_in": 15117, "tokens_out": 4179, "ctx_chars": 11062, "ctx_sha": "2b5cd4f5c3cc851b", "late_delivered": null, "t_end": 1791554248.2308843, "backend": "triton", "own_frac": 0.023, "ref_ms": 12.69097, "cand_ms": 11.8624, "endpoint": "endpoint-14", "r_wall_s": 25.86, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 3044, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00274", "task": "L1_085", "split": "calib", "reader": "claude-opus-5-5", "arm": "lessons_only", "label": "lessons_only", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L1_085|claude-opus-5-5|0", "design": "f1", "success": true, "score": 2.8723, "status": "ok", "correct": true, "speedup": 2.8723, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 179.7, "calls": 2, "tokens_in": 9641, "tokens_out": 9349, "ctx_chars": 1982, "ctx_sha": "54e16caf9bef9eec", "late_delivered": null, "t_end": 1791554249.4228094, "backend": "cuda", "own_frac": 1.0, "ref_ms": 3.75824, "cand_ms": 1.30843, "endpoint": "endpoint-17", "r_wall_s": 62.25, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 6329, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00256", "task": "L3_036", "split": "calib", "reader": "claude-opus-5-5", "arm": "frontier_gated", "label": "frontier_gated", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L3_036|claude-opus-5-5|0", "design": "f1", "success": true, "score": 4.0427, "status": "ok", "correct": true, "speedup": 4.0427, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 311.5, "calls": 3, "tokens_in": 20537, "tokens_out": 19840, "ctx_chars": 2260, "ctx_sha": "9b6af621f8f8a88b", "late_delivered": null, "t_end": 1791554252.142793, "backend": "cuda", "own_frac": 0.9997, "ref_ms": 29.81983, "cand_ms": 7.37625, "endpoint": "endpoint-19", "r_wall_s": 47.11, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 12420, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00288", "task": "L2_074", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "show_all", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_074|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.2644, "status": "ok", "correct": true, "speedup": 1.2644, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 63.5, "calls": 2, "tokens_in": 14332, "tokens_out": 2873, "ctx_chars": 10364, "ctx_sha": "19f3ac1a7ff915f6", "late_delivered": null, "t_end": 1791554255.446711, "backend": "triton", "own_frac": 0.0526, "ref_ms": 5.86288, "cand_ms": 4.63688, "endpoint": "endpoint-18", "r_wall_s": 16.03, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 1830, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00285", "task": "L2_046", "split": "calib", "reader": "claude-opus-5-5", "arm": "fg_target", "label": "fg_target", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_046|claude-opus-5-5|1", "design": "f1", "success": true, "score": 1.3964, "status": "ok", "correct": true, "speedup": 1.3964, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 95.9, "calls": 2, "tokens_in": 7615, "tokens_out": 2818, "ctx_chars": 3639, "ctx_sha": "fc6e42d43dfb4e75", "late_delivered": null, "t_end": 1791554270.1509051, "backend": "triton", "own_frac": 0.068, "ref_ms": 19.30308, "cand_ms": 13.82377, "endpoint": "endpoint-10", "r_wall_s": 37.19, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 1914, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00278", "task": "L2_024", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "late_show_all", "delivery": "late", "rep": 0, "p_arm": 0.111111, "block": "L2_024|claude-opus-5-5|0", "design": "f1", "success": false, "score": 0.2898, "status": "ok", "correct": true, "speedup": 0.2898, "infra": false, "infra_reruns": 0, "error": "ok", "wall_s": 177.7, "calls": 3, "tokens_in": 21043, "tokens_out": 8135, "ctx_chars": 10177, "ctx_sha": "cf1072ae0bde8a1b", "late_delivered": true, "t_end": 1791554272.2438645, "backend": "cuda", "own_frac": 0.9994, "ref_ms": 2.37665, "cand_ms": 8.20035, "endpoint": "endpoint-25", "r_wall_s": 22.97, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 3925, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00269", "task": "L2_073", "split": "calib", "reader": "claude-opus-5-5", "arm": "lessons_only", "label": "lessons_only", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_073|claude-opus-5-5|0", "design": "f1", "success": true, "score": 2.4884, "status": "ok", "correct": true, "speedup": 2.4884, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 239.7, "calls": 2, "tokens_in": 9873, "tokens_out": 16460, "ctx_chars": 1982, "ctx_sha": "54e16caf9bef9eec", "late_delivered": null, "t_end": 1791554279.8622, "backend": "cuda", "own_frac": 0.9925, "ref_ms": 5.64502, "cand_ms": 2.26853, "endpoint": "endpoint-26", "r_wall_s": 68.95, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 11920, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00280", "task": "L1_083", "split": "calib", "reader": "claude-opus-5-5", "arm": "frontier_gated", "label": "frontier_gated", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L1_083|claude-opus-5-5|0", "design": "f1", "success": true, "score": 2.5867, "status": "ok", "correct": true, "speedup": 2.5867, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 167.1, "calls": 2, "tokens_in": 9958, "tokens_out": 5844, "ctx_chars": 4252, "ctx_sha": "7621aa423896afd1", "late_delivered": null, "t_end": 1791554280.5612588, "backend": "cuda", "own_frac": 1.0, "ref_ms": 2.3607, "cand_ms": 0.91264, "endpoint": "endpoint-05", "r_wall_s": 79.41, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 3600, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00265", "task": "L1_077", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "late_show_all", "delivery": "late", "rep": 1, "p_arm": 0.111111, "block": "L1_077|claude-opus-5-5|1", "design": "f1", "success": true, "score": 3.0433, "status": "ok", "correct": true, "speedup": 3.0433, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 282.7, "calls": 3, "tokens_in": 26006, "tokens_out": 17214, "ctx_chars": 11175, "ctx_sha": "5e031feda794c122", "late_delivered": true, "t_end": 1791554281.3770978, "backend": "cuda", "own_frac": 0.9995, "ref_ms": 8.60277, "cand_ms": 2.8268, "endpoint": "endpoint-11", "r_wall_s": 74.64, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 9928, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "incorrect"} {"ep": "f1-00264", "task": "L3_039", "split": "calib", "reader": "claude-opus-5-5", "arm": "lessons_only", "label": "lessons_only", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L3_039|claude-opus-5-5|0", "design": "f1", "success": true, "score": 7.8993, "status": "ok", "correct": true, "speedup": 7.8993, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 295.1, "calls": 3, "tokens_in": 19593, "tokens_out": 20010, "ctx_chars": 1982, "ctx_sha": "54e16caf9bef9eec", "late_delivered": null, "t_end": 1791554291.9871862, "backend": "cuda", "own_frac": 0.9745, "ref_ms": 33.83648, "cand_ms": 4.28345, "endpoint": "endpoint-07", "r_wall_s": 39.13, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 12517, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00279", "task": "L1_078", "split": "calib", "reader": "claude-opus-5-5", "arm": "none", "label": "none", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L1_078|claude-opus-5-5|1", "design": "f1", "success": true, "score": 3.0377, "status": "ok", "correct": true, "speedup": 3.0377, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 196.0, "calls": 2, "tokens_in": 7086, "tokens_out": 9407, "ctx_chars": 0, "ctx_sha": "e3b0c44298fc1c14", "late_delivered": null, "t_end": 1791554302.4704452, "backend": "cuda", "own_frac": 1.0, "ref_ms": 35.50374, "cand_ms": 11.68752, "endpoint": "endpoint-23", "r_wall_s": 82.25, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 6967, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00286", "task": "L3_008", "split": "calib", "reader": "claude-opus-5-5", "arm": "lessons_only", "label": "lessons_only", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L3_008|claude-opus-5-5|1", "design": "f1", "success": true, "score": 1.302, "status": "ok", "correct": true, "speedup": 1.302, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 130.6, "calls": 2, "tokens_in": 8859, "tokens_out": 9094, "ctx_chars": 1982, "ctx_sha": "54e16caf9bef9eec", "late_delivered": null, "t_end": 1791554309.3493135, "backend": "triton", "own_frac": 0.0903, "ref_ms": 5.22887, "cand_ms": 4.01602, "endpoint": "endpoint-11", "r_wall_s": 25.11, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 6398, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00277", "task": "L1_077", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "show_all", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L1_077|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.7946, "status": "ok", "correct": true, "speedup": 1.7946, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 227.9, "calls": 2, "tokens_in": 16327, "tokens_out": 7756, "ctx_chars": 11175, "ctx_sha": "5e031feda794c122", "late_delivered": null, "t_end": 1791554319.2198844, "backend": "triton", "own_frac": 0.2929, "ref_ms": 8.6874, "cand_ms": 4.84085, "endpoint": "endpoint-23", "r_wall_s": 98.93, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 5718, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00284", "task": "L1_071", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_approach", "label": "no_approach", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L1_071|claude-opus-5-5|1", "design": "f1", "success": false, "score": 0.7373, "status": "ok", "correct": true, "speedup": 0.7373, "infra": false, "infra_reruns": 0, "error": "ok", "wall_s": 146.1, "calls": 2, "tokens_in": 8563, "tokens_out": 6208, "ctx_chars": 3030, "ctx_sha": "e6667ea63ccdb086", "late_delivered": null, "t_end": 1791554319.318585, "backend": "triton", "own_frac": 1.0, "ref_ms": 3.33312, "cand_ms": 4.52057, "endpoint": "endpoint-24", "r_wall_s": 61.92, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 4675, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00289", "task": "L2_052", "split": "calib", "reader": "claude-opus-5-5", "arm": "reader_relative", "label": "reader_relative", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_052|claude-opus-5-5|1", "design": "f1", "success": true, "score": 1.3816, "status": "ok", "correct": true, "speedup": 1.3816, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 122.8, "calls": 2, "tokens_in": 15528, "tokens_out": 4394, "ctx_chars": 11758, "ctx_sha": "d74bf4f221ceed35", "late_delivered": null, "t_end": 1791554323.681273, "backend": "triton", "own_frac": 0.1536, "ref_ms": 10.91041, "cand_ms": 7.89706, "endpoint": "endpoint-17", "r_wall_s": 61.48, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 2835, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00294", "task": "L1_065", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "show_all", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L1_065|claude-opus-5-5|0", "design": "f1", "success": true, "score": 2.2162, "status": "ok", "correct": true, "speedup": 2.2162, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 121.3, "calls": 2, "tokens_in": 15023, "tokens_out": 2187, "ctx_chars": 11175, "ctx_sha": "91e4a03c424736b2", "late_delivered": null, "t_end": 1791554352.865168, "backend": "triton", "own_frac": 0.0002, "ref_ms": 44.38278, "cand_ms": 20.02657, "endpoint": "endpoint-01", "r_wall_s": 55.99, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 1135, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00291", "task": "L2_052", "split": "calib", "reader": "claude-opus-5-5", "arm": "fg_target", "label": "fg_target", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_052|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.3962, "status": "ok", "correct": true, "speedup": 1.3962, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 159.5, "calls": 3, "tokens_in": 14778, "tokens_out": 6286, "ctx_chars": 3639, "ctx_sha": "fc6e42d43dfb4e75", "late_delivered": null, "t_end": 1791554369.7433653, "backend": "triton", "own_frac": 0.1383, "ref_ms": 10.96554, "cand_ms": 7.85375, "endpoint": "endpoint-18", "r_wall_s": 65.27, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "triton", "tok_out_first": 2996, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "forbidden"} {"ep": "f1-00287", "task": "L1_050", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "late_show_all", "delivery": "late", "rep": 0, "p_arm": 0.111111, "block": "L1_050|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.4161, "status": "ok", "correct": true, "speedup": 1.4161, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 183.7, "calls": 2, "tokens_in": 11879, "tokens_out": 9239, "ctx_chars": 13278, "ctx_sha": "cb78f644acd17887", "late_delivered": true, "t_end": 1791554369.8531978, "backend": "cuda", "own_frac": 0.999, "ref_ms": 5.29336, "cand_ms": 3.73802, "endpoint": "endpoint-16", "r_wall_s": 59.92, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 6654, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00282", "task": "L2_057", "split": "calib", "reader": "claude-opus-5-5", "arm": "lessons_only", "label": "lessons_only", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_057|claude-opus-5-5|1", "design": "f1", "success": true, "score": 3.7527, "status": "ok", "correct": true, "speedup": 3.7527, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 230.5, "calls": 3, "tokens_in": 15015, "tokens_out": 9751, "ctx_chars": 1982, "ctx_sha": "54e16caf9bef9eec", "late_delivered": null, "t_end": 1791554380.290412, "backend": "cuda", "own_frac": 1.0, "ref_ms": 6.3179, "cand_ms": 1.68354, "endpoint": "endpoint-26", "r_wall_s": 67.37, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 4832, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00283", "task": "L2_071", "split": "calib", "reader": "claude-opus-5-5", "arm": "reader_relative", "label": "reader_relative", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_071|claude-opus-5-5|0", "design": "f1", "success": true, "score": 2.4164, "status": "ok", "correct": true, "speedup": 2.4164, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 214.1, "calls": 3, "tokens_in": 28860, "tokens_out": 11338, "ctx_chars": 11758, "ctx_sha": "d74bf4f221ceed35", "late_delivered": null, "t_end": 1791554385.4167511, "backend": "cuda", "own_frac": 1.0, "ref_ms": 3.79382, "cand_ms": 1.57003, "endpoint": "endpoint-25", "r_wall_s": 62.25, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 6018, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00290", "task": "L2_052", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "show_all", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_052|claude-opus-5-5|1", "design": "f1", "success": true, "score": 1.3503, "status": "ok", "correct": true, "speedup": 1.3503, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 178.2, "calls": 2, "tokens_in": 15020, "tokens_out": 4169, "ctx_chars": 11132, "ctx_sha": "4caa9b6e3a0949e1", "late_delivered": null, "t_end": 1791554387.1785223, "backend": "triton", "own_frac": 0.1859, "ref_ms": 11.13317, "cand_ms": 8.24513, "endpoint": "endpoint-09", "r_wall_s": 91.9, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 2591, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00307", "task": "L2_008", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "show_all", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_008|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.0612, "status": "ok", "correct": true, "speedup": 1.0612, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 69.5, "calls": 2, "tokens_in": 14929, "tokens_out": 4202, "ctx_chars": 11062, "ctx_sha": "2b5cd4f5c3cc851b", "late_delivered": null, "t_end": 1791554388.7464828, "backend": "triton", "own_frac": 0.0351, "ref_ms": 12.66696, "cand_ms": 11.936, "endpoint": "endpoint-11", "r_wall_s": 16.81, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 3132, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00298", "task": "L2_018", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_numbers", "label": "no_numbers", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_018|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.0132, "status": "ok", "correct": true, "speedup": 1.0132, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 135.6, "calls": 3, "tokens_in": 22850, "tokens_out": 6802, "ctx_chars": 9905, "ctx_sha": "e6ee71c04b3e5f15", "late_delivered": null, "t_end": 1791554391.0002582, "backend": "triton", "own_frac": 0.0034, "ref_ms": 8.05799, "cand_ms": 7.95292, "endpoint": "endpoint-23", "r_wall_s": 29.16, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "triton", "tok_out_first": 2796, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "incorrect"} {"ep": "f1-00305", "task": "L2_085", "split": "calib", "reader": "claude-opus-5-5", "arm": "reader_relative", "label": "reader_relative", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_085|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.6007, "status": "ok", "correct": true, "speedup": 1.6007, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 92.3, "calls": 2, "tokens_in": 17179, "tokens_out": 5237, "ctx_chars": 12499, "ctx_sha": "a7d944e333fa41a4", "late_delivered": null, "t_end": 1791554394.7685447, "backend": "triton", "own_frac": 0.2825, "ref_ms": 4.71613, "cand_ms": 2.94633, "endpoint": "endpoint-01", "r_wall_s": 31.32, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 3356, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00299", "task": "L1_071", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "show_all", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L1_071|claude-opus-5-5|0", "design": "f1", "success": false, "score": 0.7404, "status": "ok", "correct": true, "speedup": 0.7404, "infra": false, "infra_reruns": 0, "error": "ok", "wall_s": 130.1, "calls": 2, "tokens_in": 15233, "tokens_out": 2620, "ctx_chars": 11175, "ctx_sha": "91e4a03c424736b2", "late_delivered": null, "t_end": 1791554400.2748334, "backend": "triton", "own_frac": 0.0008, "ref_ms": 3.33939, "cand_ms": 4.51038, "endpoint": "endpoint-24", "r_wall_s": 66.77, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 1551, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00292", "task": "L2_057", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_approach", "label": "no_approach", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_057|claude-opus-5-5|1", "design": "f1", "success": true, "score": 2.0437, "status": "ok", "correct": true, "speedup": 2.0437, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 191.2, "calls": 2, "tokens_in": 8271, "tokens_out": 7886, "ctx_chars": 3080, "ctx_sha": "7364d33fba159fed", "late_delivered": null, "t_end": 1791554407.821134, "backend": "cuda", "own_frac": 1.0, "ref_ms": 6.29579, "cand_ms": 3.0806, "endpoint": "endpoint-05", "r_wall_s": 76.74, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 5734, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00304", "task": "L1_083", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "show_all", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L1_083|claude-opus-5-5|0", "design": "f1", "success": true, "score": 2.5354, "status": "ok", "correct": true, "speedup": 2.5354, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 124.0, "calls": 2, "tokens_in": 14492, "tokens_out": 3625, "ctx_chars": 10362, "ctx_sha": "51d761e43339b5b6", "late_delivered": null, "t_end": 1791554415.9859142, "backend": "triton", "own_frac": 1.0, "ref_ms": 2.37051, "cand_ms": 0.93497, "endpoint": "endpoint-07", "r_wall_s": 55.92, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 2240, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00301", "task": "L3_004", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "show_all", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L3_004|claude-opus-5-5|0", "design": "f1", "success": true, "score": 3.4046, "status": "ok", "correct": true, "speedup": 3.4046, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 139.2, "calls": 2, "tokens_in": 13905, "tokens_out": 11456, "ctx_chars": 6513, "ctx_sha": "1aa698d3a9589f25", "late_delivered": null, "t_end": 1791554419.086752, "backend": "cuda", "own_frac": 1.0, "ref_ms": 2.30323, "cand_ms": 0.67651, "endpoint": "endpoint-16", "r_wall_s": 19.23, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 7468, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00295", "task": "L3_042", "split": "calib", "reader": "claude-opus-5-5", "arm": "reader_relative", "label": "reader_relative", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L3_042|claude-opus-5-5|1", "design": "f1", "success": true, "score": 3.8838, "status": "ok", "correct": true, "speedup": 3.8838, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 215.5, "calls": 3, "tokens_in": 29121, "tokens_out": 17455, "ctx_chars": 8756, "ctx_sha": "26dc0616aa00d951", "late_delivered": null, "t_end": 1791554463.7733033, "backend": "cuda", "own_frac": 0.8702, "ref_ms": 81.14464, "cand_ms": 20.89291, "endpoint": "endpoint-14", "r_wall_s": 29.31, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 9774, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00296", "task": "L2_018", "split": "calib", "reader": "claude-opus-5-5", "arm": "none", "label": "none", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_018|claude-opus-5-5|1", "design": "f1", "success": true, "score": 1.0101, "status": "ok", "correct": true, "speedup": 1.0101, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 216.5, "calls": 4, "tokens_in": 22164, "tokens_out": 11691, "ctx_chars": 0, "ctx_sha": "e3b0c44298fc1c14", "late_delivered": null, "t_end": 1791554465.9538867, "backend": "cuda", "own_frac": 0.0034, "ref_ms": 8.04362, "cand_ms": 7.96309, "endpoint": "endpoint-24", "r_wall_s": 40.45, "flags": [], "submit_turn": 3, "turns": 4, "tech_src": "cuda", "tok_out_first": 3944, "stops": ["end_turn", "end_turn", "end_turn", "end_turn"], "first_check": "incorrect"} {"ep": "f1-00300", "task": "L2_071", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_approach", "label": "no_approach", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_071|claude-opus-5-5|1", "design": "f1", "success": true, "score": 2.7734, "status": "ok", "correct": true, "speedup": 2.7734, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 195.4, "calls": 2, "tokens_in": 8581, "tokens_out": 9083, "ctx_chars": 3080, "ctx_sha": "7364d33fba159fed", "late_delivered": null, "t_end": 1791554467.6748881, "backend": "cuda", "own_frac": 1.0, "ref_ms": 4.01069, "cand_ms": 1.44612, "endpoint": "endpoint-11", "r_wall_s": 62.47, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 6629, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00306", "task": "L2_057", "split": "calib", "reader": "claude-opus-5-5", "arm": "show_all", "label": "late_show_all", "delivery": "late", "rep": 1, "p_arm": 0.111111, "block": "L2_057|claude-opus-5-5|1", "design": "f1", "success": true, "score": 4.0135, "status": "ok", "correct": true, "speedup": 4.0135, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 161.0, "calls": 2, "tokens_in": 10282, "tokens_out": 8405, "ctx_chars": 9855, "ctx_sha": "1a6a80cca09cee48", "late_delivered": true, "t_end": 1791554470.3783317, "backend": "cuda", "own_frac": 1.0, "ref_ms": 6.29253, "cand_ms": 1.56783, "endpoint": "endpoint-05", "r_wall_s": 60.53, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 5703, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00297", "task": "L2_069", "split": "calib", "reader": "claude-opus-5-5", "arm": "none", "label": "none", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_069|claude-opus-5-5|0", "design": "f1", "success": true, "score": 2.8356, "status": "ok", "correct": true, "speedup": 2.8356, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 250.4, "calls": 3, "tokens_in": 13650, "tokens_out": 11706, "ctx_chars": 0, "ctx_sha": "e3b0c44298fc1c14", "late_delivered": null, "t_end": 1791554502.5466511, "backend": "cuda", "own_frac": 1.0, "ref_ms": 3.80162, "cand_ms": 1.34066, "endpoint": "endpoint-18", "r_wall_s": 64.62, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 6291, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00293", "task": "L1_078", "split": "calib", "reader": "claude-opus-5-5", "arm": "reader_relative", "label": "reader_relative", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L1_078|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.9875, "status": "ok", "correct": true, "speedup": 1.9875, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 294.1, "calls": 3, "tokens_in": 28973, "tokens_out": 8916, "ctx_chars": 12197, "ctx_sha": "d2edcc521b94d748", "late_delivered": null, "t_end": 1791554512.4571643, "backend": "triton", "own_frac": 0.9998, "ref_ms": 35.87502, "cand_ms": 18.05007, "endpoint": "endpoint-19", "r_wall_s": 110.56, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "triton", "tok_out_first": 4821, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "runtime_error"} {"ep": "f1-00308", "task": "L2_008", "split": "calib", "reader": "claude-opus-5-5", "arm": "none", "label": "none", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_008|claude-opus-5-5|1", "design": "f1", "success": true, "score": 4.173, "status": "ok", "correct": true, "speedup": 4.173, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 202.0, "calls": 3, "tokens_in": 16592, "tokens_out": 15541, "ctx_chars": 0, "ctx_sha": "e3b0c44298fc1c14", "late_delivered": null, "t_end": 1791554521.3398209, "backend": "cuda", "own_frac": 0.9993, "ref_ms": 12.69189, "cand_ms": 3.04145, "endpoint": "endpoint-09", "r_wall_s": 27.63, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 8527, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00302", "task": "L2_069", "split": "calib", "reader": "claude-opus-5-5", "arm": "fg_target", "label": "fg_target", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L2_069|claude-opus-5-5|0", "design": "f1", "success": true, "score": 2.1617, "status": "ok", "correct": true, "speedup": 2.1617, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 243.7, "calls": 3, "tokens_in": 23182, "tokens_out": 11055, "ctx_chars": 7376, "ctx_sha": "cc69cebe7b058084", "late_delivered": null, "t_end": 1791554524.283968, "backend": "cuda", "own_frac": 1.0, "ref_ms": 3.62898, "cand_ms": 1.67878, "endpoint": "endpoint-01", "r_wall_s": 83.1, "flags": [], "submit_turn": 2, "turns": 3, "tech_src": "cuda", "tok_out_first": 5868, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "no_custom_kernel"} {"ep": "f1-00317", "task": "L2_067", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_numbers", "label": "no_numbers", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_067|claude-opus-5-5|1", "design": "f1", "success": false, "score": 0.0254, "status": "ok", "correct": true, "speedup": 0.0254, "infra": false, "infra_reruns": 0, "error": "ok", "wall_s": 144.1, "calls": 2, "tokens_in": 14816, "tokens_out": 4806, "ctx_chars": 11018, "ctx_sha": "17014a2bc1e52132", "late_delivered": null, "t_end": 1791554535.0973628, "backend": "triton", "own_frac": 1.0, "ref_ms": 12.89123, "cand_ms": 508.38242, "endpoint": "endpoint-26", "r_wall_s": 43.03, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 3556, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00314", "task": "L3_049", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_numbers", "label": "no_numbers", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L3_049|claude-opus-5-5|0", "design": "f1", "success": true, "score": 10.0, "status": "ok", "correct": true, "speedup": 10.9142, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 153.7, "calls": 3, "tokens_in": 16786, "tokens_out": 9532, "ctx_chars": 2425, "ctx_sha": "5c5e8811718de973", "late_delivered": null, "t_end": 1791554539.1486635, "backend": "triton", "own_frac": 0.6994, "ref_ms": 12.05247, "cand_ms": 1.10429, "endpoint": "endpoint-10", "r_wall_s": 35.17, "flags": ["excessive_speedup>10.0x: audit"], "submit_turn": 2, "turns": 3, "tech_src": "triton", "tok_out_first": 3926, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "incorrect"} {"ep": "f1-00316", "task": "L1_050", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_numbers", "label": "no_numbers", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L1_050|claude-opus-5-5|0", "design": "f1", "success": true, "score": 1.2587, "status": "ok", "correct": true, "speedup": 1.2587, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 151.4, "calls": 2, "tokens_in": 16852, "tokens_out": 5858, "ctx_chars": 13273, "ctx_sha": "0b424ce9f95f6b0e", "late_delivered": null, "t_end": 1791554540.1822321, "backend": "triton", "own_frac": 0.9918, "ref_ms": 5.29777, "cand_ms": 4.20878, "endpoint": "endpoint-25", "r_wall_s": 50.44, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 3763, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00243", "task": "L3_024", "split": "calib", "reader": "claude-opus-5-5", "arm": "frontier_gated", "label": "frontier_gated", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L3_024|claude-opus-5-5|1", "design": "f1", "success": false, "score": 0.8192, "status": "ok", "correct": true, "speedup": 0.8192, "infra": false, "infra_reruns": 0, "error": "ok", "wall_s": 674.6, "calls": 5, "tokens_in": 82247, "tokens_out": 47231, "ctx_chars": 2759, "ctx_sha": "e8b611ac667b08e3", "late_delivered": null, "t_end": 1791554543.2654245, "backend": "cuda", "own_frac": 0.0077, "ref_ms": 3.45066, "cand_ms": 4.21199, "endpoint": "endpoint-17", "r_wall_s": 86.97, "flags": [], "submit_turn": 4, "turns": 5, "tech_src": "cuda", "tok_out_first": 16952, "stops": ["end_turn", "end_turn", "end_turn", "end_turn", "end_turn"], "first_check": "runtime_error"} {"ep": "f1-00313", "task": "L1_062", "split": "calib", "reader": "claude-opus-5-5", "arm": "none", "label": "none", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L1_062|claude-opus-5-5|1", "design": "f1", "success": false, "score": 0.4794, "status": "ok", "correct": true, "speedup": 0.4794, "infra": false, "infra_reruns": 0, "error": "ok", "wall_s": 181.5, "calls": 2, "tokens_in": 7306, "tokens_out": 9548, "ctx_chars": 0, "ctx_sha": "e3b0c44298fc1c14", "late_delivered": null, "t_end": 1791554561.7527988, "backend": "cuda", "own_frac": 0.9998, "ref_ms": 9.9985, "cand_ms": 20.8561, "endpoint": "endpoint-23", "r_wall_s": 72.69, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 6787, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00318", "task": "L3_049", "split": "calib", "reader": "claude-opus-5-5", "arm": "lessons_only", "label": "lessons_only", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L3_049|claude-opus-5-5|1", "design": "f1", "success": true, "score": 10.0, "status": "ok", "correct": true, "speedup": 10.1921, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 167.0, "calls": 3, "tokens_in": 15851, "tokens_out": 9296, "ctx_chars": 1982, "ctx_sha": "54e16caf9bef9eec", "late_delivered": null, "t_end": 1791554561.806264, "backend": "triton", "own_frac": 0.7138, "ref_ms": 12.02298, "cand_ms": 1.17963, "endpoint": "endpoint-09", "r_wall_s": 21.34, "flags": ["excessive_speedup>10.0x: audit"], "submit_turn": 2, "turns": 3, "tech_src": "triton", "tok_out_first": 3302, "stops": ["end_turn", "end_turn", "end_turn"], "first_check": "incorrect"} {"ep": "f1-00310", "task": "L3_008", "split": "calib", "reader": "claude-opus-5-5", "arm": "none", "label": "none", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L3_008|claude-opus-5-5|1", "design": "f1", "success": true, "score": 1.4105, "status": "ok", "correct": true, "speedup": 1.4105, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 211.2, "calls": 2, "tokens_in": 10403, "tokens_out": 19525, "ctx_chars": 0, "ctx_sha": "e3b0c44298fc1c14", "late_delivered": null, "t_end": 1791554564.0758646, "backend": "cuda", "own_frac": 0.1826, "ref_ms": 5.25378, "cand_ms": 3.72485, "endpoint": "endpoint-16", "r_wall_s": 31.95, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 13697, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00324", "task": "L3_046", "split": "calib", "reader": "claude-opus-5-5", "arm": "lessons_only", "label": "lessons_only", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L3_046|claude-opus-5-5|0", "design": "f1", "success": true, "score": 2.1114, "status": "ok", "correct": true, "speedup": 2.1114, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 120.3, "calls": 2, "tokens_in": 9691, "tokens_out": 8585, "ctx_chars": 1982, "ctx_sha": "54e16caf9bef9eec", "late_delivered": null, "t_end": 1791554586.2990649, "backend": "triton", "own_frac": 0.4921, "ref_ms": 4.01096, "cand_ms": 1.89963, "endpoint": "endpoint-11", "r_wall_s": 27.88, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 6388, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00322", "task": "L1_062", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_approach", "label": "no_approach", "delivery": "first", "rep": 0, "p_arm": 0.111111, "block": "L1_062|claude-opus-5-5|0", "design": "f1", "success": false, "score": 0.6279, "status": "ok", "correct": true, "speedup": 0.6279, "infra": false, "infra_reruns": 0, "error": "ok", "wall_s": 169.4, "calls": 2, "tokens_in": 8955, "tokens_out": 6888, "ctx_chars": 3080, "ctx_sha": "7364d33fba159fed", "late_delivered": null, "t_end": 1791554588.465791, "backend": "triton", "own_frac": 0.9997, "ref_ms": 10.25575, "cand_ms": 16.3341, "endpoint": "endpoint-07", "r_wall_s": 71.53, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "triton", "tok_out_first": 4994, "stops": ["end_turn", "end_turn"], "first_check": "ok"} {"ep": "f1-00315", "task": "L2_027", "split": "calib", "reader": "claude-opus-5-5", "arm": "no_approach", "label": "no_approach", "delivery": "first", "rep": 1, "p_arm": 0.111111, "block": "L2_027|claude-opus-5-5|1", "design": "f1", "success": true, "score": 4.0419, "status": "ok", "correct": true, "speedup": 4.0419, "infra": false, "infra_reruns": 0, "error": "", "wall_s": 209.3, "calls": 2, "tokens_in": 9751, "tokens_out": 12684, "ctx_chars": 3080, "ctx_sha": "7364d33fba159fed", "late_delivered": null, "t_end": 1791554596.4542704, "backend": "cuda", "own_frac": 0.0007, "ref_ms": 20.11612, "cand_ms": 4.97689, "endpoint": "endpoint-17", "r_wall_s": 47.89, "flags": [], "submit_turn": 1, "turns": 2, "tech_src": "cuda", "tok_out_first": 9288, "stops": ["end_turn", "end_turn"], "first_check": "ok"}