run-0107
capability
citable URL: https://halobench.com/records/run-0107/ — this address never moves; the anchor /records/#run-0107 keeps resolving
date 2026-08-11 · window 2026-08-11T13:55:56Z → 2026-08-11T16:31:23Z
suite tau2-bench-airline@668d3bc · config cfg-0028
depth 32,768 tokens · scoring deterministic
Metrics
| metric | value |
|---|---|
| tasks_passed | 7 |
| tasks_total | 12 |
| mean_score | 0.5833 |
turns: total 567 · median 26 · p90 94 · max 201 — machine-written trailer, parsed at build time
Guard chain
not a performance run — no guard required
Energy
metered window: eng-0061
Per-task trace — the full record, no truncation
| task | reward | turns | seconds |
|---|---|---|---|
| 0 | 1 | 26 | 202 |
| 1 | 0 | 24 | 150 |
| 2 | 0 | 94 | 940 |
| 3 | 1 | 14 | 92 |
| 4 | 1 | 28 | 206 |
| 5 | 1 | 22 | 213 |
| 6 | 1 | 20 | 147 |
| 8 | 1 | 22 | 175 |
| 9 | 1 | 16 | 119 |
| 10 | 0 | 201 | 1562 |
| 11 | 0 | 62 | 410 |
| 12 | 0 | 38 | 534 |
reward null = infrastructure error, excluded from every paired computation
Cited by — computed at build time, never stored
claims clm-0047
model pages nemotron3-super