run-0271
capability
citable URL: https://halobench.com/records/run-0271/ — this address never moves; the anchor /records/#run-0271 keeps resolving
date 2026-08-15 · window 2026-08-15T02:11:33Z → 2026-08-15T05:21:33Z
suite tau2-bench-airline@668d3bc · config cfg-0066
depth 32,768 tokens · scoring deterministic
Metrics
| metric | value |
|---|---|
| tasks_passed | 15 |
| tasks_total | 26 |
| mean_score | 0.5769 |
turns: total 621 · median 24 · p90 36 · max 44 — machine-written trailer, parsed at build time
Guard chain
not a performance run — no guard required
Per-task trace — the full record, no truncation
| task | reward | turns | seconds |
|---|---|---|---|
| 0 | 1 | 12 | 149 |
| 1 | 1 | 20 | 153 |
| 2 | 1 | 24 | 576 |
| 3 | 1 | 14 | 144 |
| 4 | 1 | 18 | 165 |
| 5 | 0 | 40 | 801 |
| 6 | 1 | 16 | 228 |
| 7 | 0 | 16 | 319 |
| 8 | 1 | 30 | 259 |
| 9 | 1 | 24 | 314 |
| 10 | 1 | 26 | 507 |
| 11 | 0 | 18 | 243 |
| 12 | 1 | 22 | 357 |
| 13 | 1 | 8 | 50 |
| 14 | 1 | 34 | 724 |
| 15 | 0 | 22 | 355 |
| 16 | 0 | 19 | 240 |
| 17 | 0 | 28 | 614 |
| 18 | 0 | 24 | 237 |
| 19 | 0 | 20 | 259 |
| 20 | 1 | 32 | 392 |
| 21 | 1 | 36 | 701 |
| 22 | 0 | 24 | 488 |
| 23 | 1 | 44 | 1176 |
| 24 | 0 | 24 | 277 |
| 25 | 0 | 26 | 368 |
reward null = infrastructure error, excluded from every paired computation