← older: fast2__qwen3.6-35b-a3b__20260709-030814all runs · task boardnewer: fast__qwen3.6-27b__20260709-070108

fast2__qwen3.6-35b-a3b__20260709-033802

No notes for this run yet. Add runs/fast2__qwen3.6-35b-a3b__20260709-033802/NOTES.md — what was run & what changed, results, interpretation, next steps — and it will render here.

Run details

modelllama-local/qwen3.6-35b-a3bagentharnesses.subagents_pi:SubagentsPithinkingonreasoning budgetdirect (server/none)* — see journal for the authoritative mechanismmaxTokens / contextWindow65536 / 196608agent timeout ×2.0trials6 of 6 — 2 pass · 3 fail · 1 erroredmean reward0.33tokens (job total)54,413,594 in / 613,406 outstarted / finished2026-07-09T03:38 / 2026-07-09T07:01wall clock3h23m

Errored trials: AgentTimeoutError (fix-ocaml-gc)

Tasks

financial-document-processor — 0/1 passed

#resulttotalagentin/out tokflags
1FAIL10m39s9m46s977160/29079
claimed success but the verifier did NOT pass (heuristic)long reasoning (12,185 chars)
🔍 view

fix-ocaml-gc — 0/1 passed

#resulttotalagentin/out tokflags
1ERR2h02m2h00m23697627/397828
trial errored: AgentTimeoutErrorempty final message (no text, no tool call)long reasoning (18,772 chars) ×13loop-guard blocked a repeated call ×214
🔍 view

git-multibranch — 0/1 passed

#resulttotalagentin/out tokflags
1FAIL5m39s4m51s1774607/18600
claimed success but the verifier did NOT pass (heuristic)
🔍 view

headless-terminal — 1/1 passed

#resulttotalagentin/out tokflags
1PASS12m25s11m29s9483589/37792
long reasoning (19,832 chars) ×2loop-guard blocked a repeated callsubagent 'worker' failed
🔍 view

mailman — 0/1 passed

#resulttotalagentin/out tokflags
1FAIL46m19s44m55s18193103/120233
a bash command timed out ×3empty final message (no text, no tool call)long reasoning (34,455 chars) ×5
🔍 view

reshard-c4-data — 1/1 passed

#resulttotalagentin/out tokflags
1PASS5m49s4m09s287508/9874
🔍 view