PaperBench: Evaluating AI’s Ability to Replicate AI Research
核心要点速览
We introduce PaperBench, a benchmark evaluating the ability of AI agents to replicate state-of-the-art AI research.
查看完整报道与分享
完整图文、背景资料与讨论归原始发布平台 OpenAI News 所有。
We introduce PaperBench, a benchmark evaluating the ability of AI agents to replicate state-of-the-art AI research.
完整图文、背景资料与讨论归原始发布平台 OpenAI News 所有。