智能体
arxiv.org ·
EvoMaster：面向规模化智能体科学的基础性进化智能体框架
该 arXiv 论文提出 EvoMaster，一个面向 Agentic Science at Scale 的进化智能体框架，针对科学领域智能体基础设施重复搭建（loop fragmentation）与长周期研究中证据和经验丢失（loop discontinuity）两个问题，设计了 Execution、Exploration、Evolution 三层嵌套循环的 Loop Research 机制，使外部证据可持久化并影响后续决策。在覆盖科研、编程与推理的十个基准上，使用 GPT-5.4 的四个智能体对比中 EvoMaster 取得最高平均分 58.02%，高于最强竞品 Codex 的 40.29%，且成本低 35.6%。论文为预印本（v5，2026-09-10 修订），基准细节与评测协议需查阅全文确认。
AI推荐理由 arXiv 预印本，作者团队规模大且含 Weinan E、Siheng Chen 等知名研究者，摘要给出明确方法（E3 三层嵌套循环）与量化结果（十个基准平均 58.02%，对比 Codex 40.29%，成本低 35.6%），信息密度高。局限在于仅为预印本、未经同行评审，且摘要未披露基准构成、评测协议与成本口径，58.02% 的绝对分数偏低，实际泛化能力待验证；v5 修订说明仍在迭代。
78
正文
Computer Science > Artificial Intelligence
arXiv:2604.17406 (cs)
[Submitted on 19 Apr 2026 (v1), last revised 10 Sep 2026 (this version, v5)]
Title:EvoMaster: A Foundational Evolving Agent Framework for Agentic Science at Scale
Authors:Xinyu Zhu, Yuzhu Cai, Zexi Liu, Cheng Wang, Fengyang Li, Wenkai Jin, Wanxu Liu, Zehao Bing, Bingyang Zheng, Jingyi Chai, Shuo Tang, Rui Ye, Yuwen Du, Xianghe Pang, Yaxin Du, Tingjia Miao, Yuzhi Zhang, Ruoxue Liao, Zhaohan Ding, Linfeng Zhang, Yanfeng Wang, Weinan E, Siheng Chen
View a PDF of the paper titled EvoMaster: A Foundational Evolving Agent Framework for Agentic Science at Scale, by Xinyu Zhu and 22 other authors
View PDF
HTML (experimental)
Abstract:The convergence of large language models and agents is catalyzing a new era of scientific discovery: Agentic Science. However, common agent infrastructure is repeatedly rebuilt across scientific fields (loop fragmentation) and useful evidence and experience are lost in long-horizon research (loop discontinuity), bringing obstacles to Agentic Science at Scale. We introduce EvoMaster, a foundational evolving agent framework for Agentic Science at Scale. EvoMaster handles loop fragmentation and loop discontinuity by implementing Loop Research in which external evidence persists and improves later decisions. Through three nested loops, Execution, Exploration and Evolution (E$^3$), loop research connects research within runs, across experiments and across studies. Across ten benchmarks spanning scientific research, coding and reasoning, EvoMaster achieves the best score among four agents using GPT-5.4, reaching a mean score of 58.02%, and outperforms the strongest competing agent Codex(40.29%) while costing 35.6% less. These results show that a shared loop-research foundation can support diverse scientific agents at scale.
Comments:
Subjects:
Artificial Intelligence (cs.AI)
Cite as:
arXiv:2604.17406 [cs.AI]
(or
arXiv:2604.17406v5 [cs.AI] for this version)
https://doi.org/10.48550/arXiv.2604.17406
Focus to learn more
arXiv-issued DOI via DataCite
Submission history
From: Zhu Xinyu [view email]
[v1]
Sun, 19 Apr 2026 12:26:05 UTC (3,020 KB)
[v2]
Tue, 21 Apr 2026 06:36:45 UTC (3,020 KB)
[v3]
Sun, 7 Jun 2026 09:35:26 UTC (3,020 KB)
[v4]
Wed, 1 Jul 2026 09:10:53 UTC (3,058 KB)
[v5]
Thu, 10 Sep 2026 13:10:40 UTC (618 KB)
Full-text links:
Access Paper:
View a PDF of the paper titled EvoMaster: A Foundational Evolving Agent Framework for Agentic Science at Scale, by Xinyu Zhu and 22 other authors
View PDF
HTML (experimental)
TeX Source
view license
Current browse context:
cs.AI
new
|
recent
| 2026-04
Change to browse by:
cs
References & Citations
NASA ADS
Google Scholar
Semantic Scholar
export BibTeX citation
Loading...
BibTeX formatted citation
×
loading...
Data provided by:
Bookmark
Bibliographic Tools
Bibliographic and Citation Tools
Bibliographic Explorer Toggle
Bibliographic Explorer (What is the Explorer?)
Connected Papers Toggle
Connected Papers (What is Connected Papers?)
Litmaps Toggle
Litmaps (What is Litmaps?)
scite.ai Toggle
scite Smart Citations (What are Smart Citations?)
Code, Data, Media
Code, Data and Media Associated with this Article
alphaXiv Toggle
alphaXiv (What is alphaXiv?)
Links to Code Toggle
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub Toggle
DagsHub (What is DagsHub?)
GotitPub Toggle
Gotit.pub (What is GotitPub?)
Huggingface Toggle
Hugging Face (What is Huggingface?)
ScienceCast Toggle
ScienceCast (What is ScienceCast?)
Demos
Demos
Replicate Toggle
Replicate (What is Replicate?)
Spaces Toggle
Hugging Face Spaces (What is Spaces?)
Spaces Toggle
TXYZ.AI (What is TXYZ.AI?)
Related Papers
Recommenders and Search Tools
Link to Influence Flower
Influence Flower (What are Influence Flowers?)
Core recommender toggle
CORE Recommender (What is CORE?)
Author
Venue
Institution
Topic
About arXivLabs
arXivLabs: experimental projects with community collaborators
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Which authors of this paper are endorsers? |
Disable MathJax (What is MathJax?)