Tag
1 article
Anthropic's Claude Fable 5 outperforms OpenAI's GPT-5.5 by 13 points on the toughest FrontierMath problems, marking a significant leap in AI mathematical reasoning.
