← older: medic__qwen3.6-35b-a3b__20260710-190552all runs · task boardnewer: medic__qwen3.6-35b-a3b__20260710-201401 →

fast2__qwen3.6-35b-a3b__20260710-192053

No notes for this run yet. Add runs/fast2__qwen3.6-35b-a3b__20260710-192053/NOTES.md — what was run & what changed, results, interpretation, next steps — and it will render here.

Run details

modelllama-local/qwen3.6-35b-a3bagentharnesses.minimal_pi:MinimalPithinkingonreasoning budgetdirect (server/none)* — see journal for the authoritative mechanismmaxTokens / contextWindow65536 / 196608agent timeout ×2.0trials6 of 6 — 1 pass · 5 failmean reward0.17tokens (job total)14,657,507 in / 194,027 outstarted / finished2026-07-10T19:20 / 2026-07-10T20:12wall clock51m51s

Tasks

financial-document-processor — 0/1 passed

#resulttotalagentin/out tokflags
1FAIL15m40s14m49s7479714/80221
—
🔍 view

fix-ocaml-gc — 1/1 passed

#resulttotalagentin/out tokflags
1PASS14m25s7m43s2026230/16317
long reasoning (18,068 chars)
🔍 view

git-multibranch — 0/1 passed

#resulttotalagentin/out tokflags
1FAIL3m01s2m02s571046/14526
—
🔍 view

headless-terminal — 0/1 passed

#resulttotalagentin/out tokflags
1FAIL4m26s3m42s796725/17548
a bash command timed out
🔍 view

mailman — 0/1 passed

#resulttotalagentin/out tokflags
1FAIL6m12s4m47s2531118/29776
claimed success but the verifier did NOT pass (heuristic)long reasoning (30,358 chars) ×2
🔍 view

reshard-c4-data — 0/1 passed

#resulttotalagentin/out tokflags
1FAIL8m05s6m17s1252674/35639
fast-timeout cut at 5mlong reasoning (19,828 chars) ×2
🔍 view