← older: suite__qwen3.6-35b-a3b__20260724-160500all runs · task boardnewer: smoke__qwen3.6-35b-a3b__20260726-025257 →

smokeqwen3.6-35b-a3b20260726-002616

No notes for this run yet. Add runs/smoke__qwen3.6-35b-a3b__20260726-002616/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 ×2.0trials9 of 9 — 8 pass · 1 failmean reward0.89tokens (job total)42,005,349 in / 630,557 outstarted / finished2026-07-26T00:26 / 2026-07-26T01:44wall clock1h17m

Tasks

cobol-modernization — 3/3 passed

#	result	total	agent	in/out tok	flags
1	PASS	4m18s	3m44s	1434867/36025	long reasoning (17,621 chars) ×3	🔍 view
2	PASS	2m56s	2m22s	1045025/23307	—	🔍 view
3	PASS	5m05s	4m32s	1302685/44651	long reasoning (12,674 chars)	🔍 view

largest-eigenval — 2/3 passed

#	result	total	agent	in/out tok	flags
1	FAIL	22m36s	22m03s	23635827/164976	long reasoning (30,499 chars)	🔍 view
2	PASS	5m50s	5m19s	6007763/35272	—	🔍 view
3	PASS	11m55s	11m24s	4494116/97750	long reasoning (20,314 chars) ×6	🔍 view

vulnerable-secret — 3/3 passed

#	result	total	agent	in/out tok	flags
1	PASS	10m16s	9m42s	1162797/94097	long reasoning (14,464 chars) ×7	🔍 view
2	PASS	6m28s	5m55s	786239/59189	long reasoning (16,249 chars) ×4	🔍 view
3	PASS	8m25s	7m51s	2136030/75290	long reasoning (13,200 chars) ×3	🔍 view