跳到正文
原文
OpenAI News·· 2025-04-02AI 评分45

PaperBench:评估AI复制AI研究能力的基准

PaperBench: Evaluating AI’s Ability to Replicate AI Research

AI 导读

OpenAI推出PaperBench基准,用于评估AI代理复制最先进AI研究的能力。

原文

We introduce PaperBench, a benchmark evaluating the ability of AI agents to replicate state-of-the-art AI research.

来源:OpenAI News · openai.com