arXiv cs.AI· Taiming Lu, Sachin Konan, Zhuang Liu·· 12 小时前AI 评分66
Scientist World：为 AI-Scientist 构建可执行的研究世界
Scientist World: Grow a Research World for AI-Scientists
AI 导读
研究团队提出 Scientist World，把每篇研究论文转成可执行环境，并发布首批 1000 个经验证的环境。在该环境中评测时，前沿 AI-Scientist 有 74% 的情况超越人类研究者；团队还记录了 6400 小时研究轨迹，并称可将小模型的研究构思能力提升 158%，使其追上前沿模型。
正文 · 原文
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.
| Subjects: | Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Machine Learning (cs.LG) |
| Cite as: | arXiv:2611.08478 [cs.AI] |
| (or arXiv:2611.08478v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2611.08478 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: TaiMing Lu [view email]
[v1]
Thu, 5 Nov 2026 18:58:17 UTC (12,169 KB)
来源：arXiv cs.AI · arxiv.org