← older: fast__qwen3.6-27b__20260706-124744all runs · task boardnewer: fast__qwen3.6-35b-a3b__20260706-171844

fast__qwen3.6-27b__20260706-132725

No notes for this run yet. Add runs/fast__qwen3.6-27b__20260706-132725/NOTES.md — what was run & what changed, results, interpretation, next steps — and it will render here.

Run details

modelllama-local/qwen3.6-27bagentharnesses.minimal_pi:MinimalPithinkingonreasoning budgetdirect (server/none)* — see journal for the authoritative mechanismmaxTokens / contextWindow65536 / 196608agent timeout ×2.0trials18 of 18 — 13 pass · 5 failmean reward0.72tokens (job total)1,790,569 in / 76,683 outstarted / finished2026-07-06T13:27 / 2026-07-06T14:50wall clock1h22m

Tasks

cancel-async-tasks — 0/2 passed

#resulttotalagentin/out tokflags
1FAIL1m22s31s16911/1669
🔍 view
2FAIL1m23s33s24918/1687
🔍 view

filter-js-from-html — 0/2 passed

#resulttotalagentin/out tokflags
1FAIL10m55s4m00s24196/10005
fast-timeout cut at 4mlong reasoning (28,588 chars)
🔍 view
2FAIL4m38s4m00s0/0
fast-timeout cut at 4m
🔍 view

fix-git — 2/2 passed

#resulttotalagentin/out tokflags
1PASS1m04s26s31821/1290
🔍 view
2PASS1m07s30s33667/1485
🔍 view

nginx-request-logging — 2/2 passed

#resulttotalagentin/out tokflags
1PASS1m17s39s34401/2024
🔍 view
2PASS1m20s40s49242/2161
🔍 view

openssl-selfsigned-cert — 2/2 passed

#resulttotalagentin/out tokflags
1PASS1m11s35s24393/1892
🔍 view
2PASS1m31s54s40061/2892
🔍 view

query-optimize — 1/2 passed

#resulttotalagentin/out tokflags
1FAIL10m31s3m45s56350/3064
🔍 view
2PASS16m35s10m00s44599/2883
fast-timeout cut at 10m
🔍 view

regex-log — 2/2 passed

#resulttotalagentin/out tokflags
1PASS10m44s10m00s1951/8328
fast-timeout cut at 10mlong reasoning (20,274 chars)
🔍 view
2PASS3m38s2m55s34945/9237
empty final message (no text, no tool call)long reasoning (20,240 chars)
🔍 view

sanitize-git-repo — 2/2 passed

#resulttotalagentin/out tokflags
1PASS2m33s1m55s287608/3713
🔍 view
2PASS3m37s2m59s629419/4173
🔍 view

sparql-university — 2/2 passed

#resulttotalagentin/out tokflags
1PASS4m37s3m52s289706/10632
long reasoning (26,050 chars)
🔍 view
2PASS4m45s4m00s166381/9548
fast-timeout cut at 4mlong reasoning (25,276 chars)
🔍 view