run-0433
capability
citable URL: https://halobench.com/records/run-0433/ — this address never moves; the anchor /records/#run-0433 keeps resolving
date 2026-08-17 · window 2026-08-17T19:09:14Z → 2026-08-17T19:48:02Z
suite tau2-bench-airline@668d3bc · config cfg-0111
depth 32,768 tokens · scoring deterministic
Metrics
| metric | value |
|---|---|
| tasks_passed | 14 |
| tasks_total | 26 |
| mean_score | 0.5385 |
turns: total 780 · median 30 · p90 44 · max 46 — machine-written trailer, parsed at build time
Guard chain
guarded by run-0395
Energy
metered window: eng-0168
Per-task trace — the full record, no truncation
| task | reward | turns | seconds |
|---|---|---|---|
| 0 | 0 | 16 | 44 |
| 1 | 1 | 32 | 60 |
| 2 | 1 | 30 | 87 |
| 3 | 1 | 14 | 30 |
| 4 | 1 | 26 | 47 |
| 5 | 1 | 24 | 56 |
| 6 | 1 | 32 | 138 |
| 7 | 0 | 44 | 98 |
| 8 | 0 | 28 | 93 |
| 9 | 1 | 22 | 54 |
| 10 | 0 | 32 | 112 |
| 11 | 0 | 44 | 126 |
| 12 | 1 | 22 | 59 |
| 13 | 1 | 14 | 42 |
| 14 | 0 | 40 | 129 |
| 15 | 0 | 28 | 80 |
| 16 | 1 | 28 | 86 |
| 17 | 1 | 34 | 158 |
| 18 | 0 | 34 | 82 |
| 19 | 1 | 16 | 41 |
| 20 | 0 | 28 | 87 |
| 21 | 0 | 46 | 152 |
| 22 | 1 | 40 | 92 |
| 23 | 0 | 38 | 144 |
| 24 | 0 | 40 | 116 |
| 25 | 1 | 28 | 97 |
reward null = infrastructure error, excluded from every paired computation
Cited by — computed at build time, never stored
claims clm-0082
model pages qwen3-coder-next-80b