DiscoveryRSI-27B 框架发布：假设对应率 6.3% 超越 GPT-6 和 Claude Opus 5
科学发现 RSI 要可证伪假设树：多校 DiscoveryRSI，假设对应率 6.3% vs GPT-6 2.5%
科学发现要学的是「有价值且可检验的假设」，不是再堆一篇综述式生成。
DiscoveryRSI-27B 框架发布 (arXiv:2610.23839)
“Authors: Jianyuan Zhong , Chenwei Wu , Zhaochen Yu , Xixian Jiang , Xuewei Li , Hao Xiao , Xilin Dai , Junyang Wang , Jiaru Zou , Zhiqing Cu”
Authors:Jianyuan Zhong, Chenwei Wu, Zhaochen Yu, Xixian Jiang, Xuewei Li, Hao Xiao, Xilin Dai, Junyang Wang, Jiaru Zou, Zhiqing Cui, Boyuan Zheng, Jing Xiong, Zhen Zhang, Qiang Gao, Michael Morley, Luis Nakayama, Qiang Xu, Liyue Shen, Tao Gui, Leo Anthony Celi, Zhenfei Yin, Yingcheng Wu, Ling Yang
Abstract:Scientific discovery requires developing worthwhile hypotheses into practical investigations, yet learning from this evolving research process remains challenging. We introduce an operational definition of valuable and testable research hypotheses: evidence-grounded claims about consequential uncertainties, coupled with feasible investigations, explicit decision criteria, and alternative explanations. Our DiscoveryRSI framework develops hypothesis trees through multi-agent evolutionary search, compiles selected investigations into executable Harbor tasks, and recursively learns through three mechanisms: hypothesis credit optimization attributes scientific assessments to documented research contributions; harness replay identifies productive strategies and recurring failures to refine research workflows; and rediscovery applies accepted improvements to new investigations, generating further experience for subsequent learning. We also introduce DFM-Bench, which provides comprehensive assessment of hypothesis quality and cumulative research progress at the levels of hypotheses, revisions, development paths and research runs. Across four life-science domains, automated evaluation shows that hypothesis quality rises from 4.86 for the base model to 5.39 after two rounds of self-improvement, with the large gains in falsifiability and data resolvability. We further validate hypothesis quality through expert review, comparing our DiscoveryRSI-27B with other research agents. Preliminary human assessments show that DiscoveryRSI-27B scores higher than GPT-6 Luna on hypothesis importance and whether claims are plausible and evidence-bounded. DiscoveryRSI-27B achieved a 6.3% rate of hypothesis correspondence with research published after the model cutoff, largely outperform GPT-6 with 2.5% and Claude Opus 5 with 2.3%, while delivering greater human-rated improvement along research paths.
| Comments: | Code: this https URL |
| Subjects: | Computation and Language (cs.CL) |
| Cite as: | arXiv:2610.23839 [cs.CL] |
| (or arXiv:2610.23839v2 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2610.23839 arXiv-issued DOI via DataCite |
Submission history
From: Ling Yang [view email]
[v1]
Fri, 16 Oct 2026 17:56:35 UTC (6,630 KB)
[v2]
Tue, 20 Oct 2026 08:05:48 UTC (6,635 KB)
来源：Agent Control Flash News · arxiv.org
关联产品与深度拆解
Cursor Agent 架构全景拆解：从影子工作区到多文件推演引擎
全面剖析 Cursor 在 Agent 模式下的底层技术实现，解析影子分支隔离、LSP 上下文剪枝以及预测性补全的核心架构细节与工程实践经验。
Vercel v0 推出全新多 Agent 协作画布与实时全栈部署预览
动态速评：v0 正在从单一组件生成器演进为全栈应用搭建协作台，尤其是与 Neon/Postgres 的直连极大降低了全栈原型验证成本。