GPT-5.5 – No ARC-AGI-3 scores

25 pointsAG2517d ago3 comments

Did the model perform poorly and OpenAI decided to not publish arc agi 3 scores? This is honestly the best benchmark right now to measure true intelligence.

3 comments

casey217d ago

ARC-AGI-3 scoring is really weird, in some views it's already saturated in others it's near 0. But I assume, since the entire benchmark IMO is a PR tool for OpenAI they will publish it eventually.

LouisLau17d ago

ARC-AGI-3 is interesting, but it’s not necessarily aligned with what models are being optimized for right now (reasoning + usefulness in real tasks).

j / k navigate · click thread line to collapse

3 comments

casey217d ago

ARC-AGI-3 scoring is really weird, in some views it's already saturated in others it's near 0. But I assume, since the entire benchmark IMO is a PR tool for OpenAI they will publish it eventually.

LouisLau17d ago

ARC-AGI-3 is interesting, but it’s not necessarily aligned with what models are being optimized for right now (reasoning + usefulness in real tasks).

j / k navigate · click thread line to collapse