AI 뉴스로 돌아가기

PaperBench: Evaluating AI’s Ability to Replicate AI Research

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

2025.04.0210원문 보기
We introduce PaperBench, a benchmark evaluating the ability of AI agents to replicate state-of-the-art AI research.