Scientist World：为 AI-Scientists 培育一个研究世界
Scientist World: Grow a Research World for AI-Scientists
Taiming Lu · Sachin Konan · Zhuang Liu
中文摘要
AI-Scientists 已经出现，但可供其成长的世界尚未建立。人类科学家将自身的研究世界视为理所当然——几个世纪积累的开放问题与共享思想静候于科学文献之中。对 AI-Scientists 而言，这个世界可读却不可运行，因此我们为其构建 Scientist World：每篇研究论文都成为一个可执行环境。自主流水线提取每篇论文的开放式研究问题，并将其提出的方法复现为基线；进入环境的 AI-Scientist 将对照人类作者所取得的成果接受评估。我们发布了首批 1,000 个经过验证的环境。以该世界作为评估基准，我们发现前沿 AI-Scientists 已有 74% 的情况下超越人类研究者。作为 AI 如何做科学的记录，它包含 6,400 小时的研究轨迹，涵盖每一个想法、实验、失败与成功。作为训练场，我们能够将一个小型模型提升（+158%），使其在研究构思上匹配前沿模型。AI-Scientists 自己的世界终于到来，它从科学本身生长而来，并与科学共同成长。
关键要点
- 01AI-Scientists 缺少可运行的研究环境，现有文献可读但不可执行
- 02将每篇论文转为可执行环境，提取开放问题并复现方法作为基线
- 03发布首批 1,000 个验证环境，前沿 AI-Scientists 在 74% 情况下超越人类作者
- 04记录 6,400 小时研究轨迹，含想法、实验、失败与成功全过程
- 05作为训练场可将小模型研究构思能力提升 158%，匹配前沿模型
解读
尚无解读。
原始英文摘要
arXiv:2611.08478v1 Announce Type: cross Abstract: AI-Scientists arrived, but a world for them to grow in did not. Human scientists take theirs for granted $-$ centuries of open problems and shared ideas, waiting in the scientific literature. To AI-Scientists, that world is readable but not runnable, so we build Scientist World for them: every research paper becomes an executable environment. Our autonomous pipeline extracts each paper's open-ended research question and reproduces its proposed method as the baseline; an AI-Scientist entering the environment is evaluated against what the human authors achieved. We release the first 1,000 verified environments. With the world as an evaluation, we found that frontier AI-Scientists can already surpass the human researchers 74% of the time. As a record of how AI does science, it contains 6,400 hours of research traces, with every idea, experiment, failure, and success. As a training ground, we are able to raise (+158%) a small model to match frontier models in research ideation. At last, AI-Scientists' own world has arrived, grown from and growing with science itself.