Somehow regresses on SWE bench?

lkbm 5 hours ago | parent | next [-]

I don't know how these benchmarks work (do you do a hundred runs? A thousand runs?), but 0.1% seems like noise.

SubiculumCode 5 hours ago | parent | prev | next [-]

That benchmark is pretty saturated, tbh. A "regression" of such small magnitude could mean many different things or nothing at all.

usaar333 5 hours ago | parent | prev [-]

i'd interpret that as rounding error. that is unchanged

swe-bench seems really hard once you are above 80%

it's not a great benchmark anymore... starting with it being python / django primarily... the industry should move to something more representative

	▲	usaar333 5 hours ago \| parent [-]
		Openai has; they don't even mention score on gpt-5.3-codex. On the other hand, it is their own verified benchmark, which is telling.