run-0287
capability
citable URL: https://halobench.com/records/run-0287/ — this address never moves; the anchor /records/#run-0287 keeps resolving
date 2026-08-15 · window 2026-08-15T22:06:31Z → 2026-08-16T01:13:01Z
suite tau2-bench-airline@668d3bc · config cfg-0084
depth 32,768 tokens · scoring deterministic
Metrics
| metric | value |
|---|---|
| tasks_passed | 22 |
| tasks_total | 26 |
| mean_score | 0.8462 |
Guard chain
waiver — This IS the capability-establishing run (kind: capability); schema only requires a guard reference on kind: performance runs. The 2026-08-15 screen's own GUARD (4/4, no cliff) ran on ROCm/f16/c32768/UD-IQ3_XXS — same fingerprint as this arm except --parallel (hazard-class, never inherited): this run supplies its own evidence instead (272 tool-call messages across 26 tasks, 0 empty assistant turns, 0 infrastructure errors — the SMOKE validity gate's own signal that the agent genuinely acted throughout).
Energy
metered window: eng-0107
Per-task trace — the full record, no truncation
| task | reward | turns | seconds |
|---|---|---|---|
| 0 | 1 | 18 | 185 |
| 1 | 1 | 24 | 223 |
| 2 | 1 | 27 | 277 |
| 3 | 0 | 16 | 104 |
| 4 | 1 | 23 | 348 |
| 5 | 1 | 24 | 268 |
| 6 | 1 | 17 | 119 |
| 7 | 0 | 48 | 496 |
| 8 | 1 | 28 | 287 |
| 9 | 1 | 37 | 409 |
| 10 | 1 | 38 | 931 |
| 11 | 1 | 21 | 256 |
| 12 | 1 | 22 | 314 |
| 13 | 1 | 8 | 58 |
| 14 | 1 | 32 | 665 |
| 15 | 1 | 22 | 340 |
| 16 | 1 | 20 | 308 |
| 17 | 1 | 27 | 743 |
| 18 | 1 | 49 | 826 |
| 19 | 1 | 18 | 140 |
| 20 | 1 | 22 | 270 |
| 21 | 0 | 30 | 550 |
| 22 | 1 | 29 | 641 |
| 23 | 0 | 46 | 1302 |
| 24 | 1 | 46 | 490 |
| 25 | 1 | 22 | 262 |
reward null = infrastructure error, excluded from every paired computation
Cited by — computed at build time, never stored
claims clm-0061
model pages deepseek-v4-flash