← older: fast__qwen3.6-35b-a3b__20260706-113104all runs · task boardnewer: fast__qwen3.6-27b__20260706-132725 →

fastqwen3.6-27b20260706-124744

No notes for this run yet. Add runs/fast__qwen3.6-27b__20260706-124744/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 ×2.0trials9 of 9 — 7 pass · 2 failmean reward0.78tokens (job total)1,240,631 in / 57,038 outstarted / finished2026-07-06T12:47 / 2026-07-06T13:27wall clock39m17s

Tasks

cancel-async-tasks — 0/1 passed

#	result	total	agent	in/out tok	flags
1	FAIL	2m09s	1m20s	42061/4154	claimed success but the verifier did NOT pass (heuristic)	🔍 view

filter-js-from-html — 0/1 passed

#	result	total	agent	in/out tok	flags
1	FAIL	5m53s	2m43s	209202/8619	—	🔍 view

fix-git — 1/1 passed

#	result	total	agent	in/out tok	flags
1	PASS	1m08s	30s	38769/1564	—	🔍 view

nginx-request-logging — 1/1 passed

#	result	total	agent	in/out tok	flags
1	PASS	1m30s	53s	55566/2711	—	🔍 view

openssl-selfsigned-cert — 1/1 passed

#	result	total	agent	in/out tok	flags
1	PASS	1m41s	1m04s	60088/3452	—	🔍 view

query-optimize — 1/1 passed

#	result	total	agent	in/out tok	flags
1	PASS	12m56s	6m19s	57068/2345	—	🔍 view

regex-log — 1/1 passed

#	result	total	agent	in/out tok	flags
1	PASS	7m22s	6m38s	325947/20739	long reasoning (21,598 chars)	🔍 view

sanitize-git-repo — 1/1 passed

#	result	total	agent	in/out tok	flags
1	PASS	3m02s	2m25s	403088/4413	—	🔍 view

sparql-university — 1/1 passed

#	result	total	agent	in/out tok	flags
1	PASS	3m32s	2m48s	48842/9041	empty final message (no text, no tool call)long reasoning (26,596 chars)	🔍 view