OpenAI’s GPT-6 Astra Posts 63% on ARC-AGI-3 Standard Evaluation
Sign in✓ Signed in
OpenAI’s GPT-6 Astra scored 63% on ARC-AGI-3 under a standard harness, while using fewer actions than the median tested human on 96% of levels. The result highlights both a major advance in interactive reasoning and the growing influence of the software layer used to evaluate agentic models.
Continue reading
The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.
Already purchased? Sign in✓ Signed in — this article isn’t included in your current plan.
Unlocking the full article...