run-0110
capability
citable URL: https://halobench.com/records/run-0110/ — this address never moves; the anchor /records/#run-0110 keeps resolving
date 2026-08-11 · window 2026-08-11T19:29:50Z → 2026-08-11T19:58:14Z
suite tau2-bench-airline@668d3bc · config cfg-0025
depth 32,768 tokens · scoring deterministic
Metrics
| metric | value |
|---|---|
| tasks_passed | 7 |
| tasks_total | 9 |
| mean_score | 0.7778 |
turns: total 230 · median 24 · p90 52 · max 52 — machine-written trailer, parsed at build time
Guard chain
not a performance run — no guard required
Energy
metered window: eng-0064
Per-task trace — the full record, no truncation
| task | reward | turns | seconds |
|---|---|---|---|
| 0 | 1 | 18 | 105 |
| 1 | 1 | 28 | 140 |
| 2 | 1 | 26 | 132 |
| 3 | 1 | 14 | 77 |
| 4 | 1 | 24 | 168 |
| 5 | 0 | 52 | 590 |
| 6 | 1 | 18 | 97 |
| 8 | 1 | 30 | 197 |
| 9 | 0 | 20 | 125 |
reward null = infrastructure error, excluded from every paired computation
Cited by — computed at build time, never stored
claims clm-0048