https://github.com/CharlesQ9/Self-Evolving-Agents

\n","updatedAt":"2025-07-29T06:36:58.491Z","author":{"_id":"65271df5abd7795aaa1fd86e","avatarUrl":"/avatars/99e1aabaefcf5e7f438817375cd1ddef.svg","fullname":"Jiahao Qiu","name":"jiahaoq","type":"user","isPro":false,"isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":5,"isUserFollowing":false}},"numEdits":0,"identifiedLanguage":{"language":"en","probability":0.8617885708808899},"editors":["jiahaoq"],"editorAvatarUrls":["/avatars/99e1aabaefcf5e7f438817375cd1ddef.svg"],"reactions":[{"reaction":"🤯","users":["XiangJinYu","tianbaoxiexxx","Merlin-Hongru"],"count":3}],"isReport":false}},{"id":"6888c535b38a417dd48bfe0b","createdAt":"2025-07-29T12:57:25.000Z","type":"comment","data":{"edited":false,"hidden":false,"latest":{"raw":"https://github.com/tarikkaya/aix","html":"

\n","updatedAt":"2025-07-29T12:57:25.255Z"},"numEdits":0,"identifiedLanguage":{"language":"en","probability":0.5513794422149658},"editors":["deleted"],"editorAvatarUrls":["deleted"],"reactions":[],"isReport":false}},{"id":"688976fa0702b1d45f250626","author":{"_id":"63d3e0e8ff1384ce6c5dd17d","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/1674830754237-63d3e0e8ff1384ce6c5dd17d.jpeg","fullname":"Librarian Bot (Bot)","name":"librarian-bot","type":"user","isPro":false,"isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":318,"isUserFollowing":false},"createdAt":"2025-07-30T01:35:54.000Z","type":"comment","data":{"edited":false,"hidden":false,"latest":{"raw":"This is an automated message from the [Librarian Bot](https://huggingface.co/librarian-bots). I found the following papers similar to this paper. \n\nThe following papers were recommended by the Semantic Scholar API \n\n* [Adaptability in Multi-Agent Reinforcement Learning: A Framework and Unified Review](https://huggingface.co/papers/2507.10142) (2025)\n* [Deep Research Agents: A Systematic Examination And Roadmap](https://huggingface.co/papers/2506.18096) (2025)\n* [From Web Search towards Agentic Deep Research: Incentivizing Search with Reasoning Agents](https://huggingface.co/papers/2506.18959) (2025)\n* [Multi-Agent Language Models: Advancing Cooperation, Coordination, and Adaptation](https://huggingface.co/papers/2506.09331) (2025)\n* [Evolutionary Perspectives on the Evaluation of LLM-Based AI Agents: A Comprehensive Survey](https://huggingface.co/papers/2506.11102) (2025)\n* [Intelligent Design 4.0: Paradigm Evolution Toward the Agentic AI Era](https://huggingface.co/papers/2506.09755) (2025)\n* [Learn as Individuals, Evolve as a Team: Multi-agent LLMs Adaptation in Embodied Environments](https://huggingface.co/papers/2506.07232) (2025)\n\n\n Please give a thumbs up to this comment if you found it helpful!\n\n If you want recommendations for any Paper on Hugging Face checkout [this](https://huggingface.co/spaces/librarian-bots/recommend_similar_papers) Space\n\n You can directly ask Librarian Bot for paper recommendations by tagging it in a comment: `@librarian-bot recommend`","html":"

This is an automated message from the Librarian Bot. I found the following papers similar to this paper.

The following papers were recommended by the Semantic Scholar API

Please give a thumbs up to this comment if you found it helpful!

If you want recommendations for any Paper on Hugging Face checkout this Space

You can directly ask Librarian Bot for paper recommendations by tagging it in a comment: \n\n@librarian-bot\n\t recommend

\n","updatedAt":"2025-07-30T01:35:54.273Z","author":{"_id":"63d3e0e8ff1384ce6c5dd17d","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/1674830754237-63d3e0e8ff1384ce6c5dd17d.jpeg","fullname":"Librarian Bot (Bot)","name":"librarian-bot","type":"user","isPro":false,"isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":318,"isUserFollowing":false}},"numEdits":0,"identifiedLanguage":{"language":"en","probability":0.7529680132865906},"editors":["librarian-bot"],"editorAvatarUrls":["https://cdn-avatars.huggingface.co/v1/production/uploads/1674830754237-63d3e0e8ff1384ce6c5dd17d.jpeg"],"reactions":[],"isReport":false}},{"id":"688a0a4b1fc00d2997b81862","author":{"_id":"65d9fc2a0e6ad24551d87a1e","avatarUrl":"/avatars/3aedb9522cc3cd08349d654f523fd792.svg","fullname":"Grant Singleton","name":"grantsing","type":"user","isPro":false,"isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":4,"isUserFollowing":false},"createdAt":"2025-07-30T12:04:27.000Z","type":"comment","data":{"edited":false,"hidden":false,"latest":{"raw":"arXiv explained breakdown of this paper 👉 https://arxivexplained.com/papers/a-survey-of-self-evolving-agents-on-path-to-artificial-super-intelligence.","html":"

arXiv explained breakdown of this paper 👉 https://arxivexplained.com/papers/a-survey-of-self-evolving-agents-on-path-to-artificial-super-intelligence.

\n","updatedAt":"2025-07-30T12:04:27.617Z","author":{"_id":"65d9fc2a0e6ad24551d87a1e","avatarUrl":"/avatars/3aedb9522cc3cd08349d654f523fd792.svg","fullname":"Grant Singleton","name":"grantsing","type":"user","isPro":false,"isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":4,"isUserFollowing":false}},"numEdits":0,"identifiedLanguage":{"language":"en","probability":0.8470942974090576},"editors":["grantsing"],"editorAvatarUrls":["/avatars/3aedb9522cc3cd08349d654f523fd792.svg"],"reactions":[],"isReport":false}}],"primaryEmailConfirmed":false,"paper":{"id":"2507.21046","authors":[{"_id":"68886bb5af872d625c10c6ba","user":{"_id":"64b1303bf460afaefcf922c2","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/noauth/0zhdXXravx0CKJLTSVDwU.jpeg","isPro":false,"fullname":"Huan-ang Gao","user":"c7w","type":"user"},"name":"Huan-ang Gao","status":"claimed_verified","statusLastChangedAt":"2025-08-26T09:51:00.700Z","hidden":false},{"_id":"68886bb5af872d625c10c6bb","name":"Jiayi Geng","hidden":false},{"_id":"68886bb5af872d625c10c6bc","user":{"_id":"639a25aba2b0b1c9d85a51e8","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/639a25aba2b0b1c9d85a51e8/pphz-MK62hPNbBkMHAkeR.jpeg","isPro":false,"fullname":"Wenyue Hua","user":"wenyueH","type":"user"},"name":"Wenyue Hua","status":"claimed_verified","statusLastChangedAt":"2025-08-06T19:32:56.971Z","hidden":false},{"_id":"68886bb5af872d625c10c6bd","name":"Mengkang Hu","hidden":false},{"_id":"68886bb5af872d625c10c6be","name":"Xinzhe Juan","hidden":false},{"_id":"68886bb5af872d625c10c6bf","user":{"_id":"6632160088f75d987d1a156f","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/6632160088f75d987d1a156f/mYlMQfK1BGWeEbOSMmeSb.jpeg","isPro":false,"fullname":"Hongzhang Liu","user":"Alphamasterliu","type":"user"},"name":"Hongzhang Liu","status":"claimed_verified","statusLastChangedAt":"2025-07-29T07:07:22.183Z","hidden":false},{"_id":"68886bb5af872d625c10c6c0","name":"Shilong Liu","hidden":false},{"_id":"68886bb5af872d625c10c6c1","name":"Jiahao Qiu","hidden":false},{"_id":"68886bb5af872d625c10c6c2","name":"Xuan Qi","hidden":false},{"_id":"68886bb5af872d625c10c6c3","name":"Yiran Wu","hidden":false},{"_id":"68886bb5af872d625c10c6c4","name":"Hongru Wang","hidden":false},{"_id":"68886bb5af872d625c10c6c5","user":{"_id":"666aa99cd1652853e4f9a8b9","avatarUrl":"/avatars/7cd5a0c34b5ccb8eff5a353d88d15a93.svg","isPro":false,"fullname":"HanXiao","user":"HanXiao1999","type":"user"},"name":"Han Xiao","status":"claimed_verified","statusLastChangedAt":"2025-07-29T12:50:59.442Z","hidden":false},{"_id":"68886bb5af872d625c10c6c6","name":"Yuhang Zhou","hidden":false},{"_id":"68886bb5af872d625c10c6c7","name":"Shaokun Zhang","hidden":false},{"_id":"68886bb5af872d625c10c6c8","user":{"_id":"65f40e83653c231cbaf7defe","avatarUrl":"/avatars/afa5ce72324112739e539865c9aee26b.svg","isPro":false,"fullname":"Jiayi Zhang","user":"didiforhugface","type":"user"},"name":"Jiayi Zhang","status":"claimed_verified","statusLastChangedAt":"2025-07-29T07:07:24.616Z","hidden":false},{"_id":"68886bb5af872d625c10c6c9","user":{"_id":"649ea7106282cb41e77760bc","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/649ea7106282cb41e77760bc/HlWjaqxr03ob93vdKg_LQ.jpeg","isPro":false,"fullname":"Isaac","user":"XiangJinYu","type":"user"},"name":"Jinyu Xiang","status":"claimed_verified","statusLastChangedAt":"2025-07-29T07:07:19.166Z","hidden":false},{"_id":"68886bb5af872d625c10c6ca","name":"Yixiong Fang","hidden":false},{"_id":"68886bb5af872d625c10c6cb","name":"Qiwen Zhao","hidden":false},{"_id":"68886bb5af872d625c10c6cc","user":{"_id":"6621f4eb64e84619e578aad6","avatarUrl":"/avatars/b1ad96ee354b999fcafb2998a636609c.svg","isPro":false,"fullname":"Dongrui Liu","user":"shenqiorient","type":"user"},"name":"Dongrui Liu","status":"claimed_verified","statusLastChangedAt":"2026-02-20T08:39:22.792Z","hidden":false},{"_id":"68886bb5af872d625c10c6cd","user":{"_id":"66e2624a436a1798365e4581","avatarUrl":"/avatars/6c605807d34faa8fb505e135a4b47776.svg","isPro":false,"fullname":"Qihan Ren","user":"jasonrqh","type":"user"},"name":"Qihan Ren","status":"claimed_verified","statusLastChangedAt":"2025-07-29T14:37:45.288Z","hidden":false},{"_id":"68886bb5af872d625c10c6ce","user":{"_id":"665e121c6007027038fd4005","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/no-auth/sIVBJAGM-Kneq9KMf8aXb.png","isPro":false,"fullname":"Cheng Qian","user":"chengq9","type":"user"},"name":"Cheng Qian","status":"claimed_verified","statusLastChangedAt":"2025-09-29T06:06:17.947Z","hidden":false},{"_id":"68886bb5af872d625c10c6cf","name":"Zhenghailong Wang","hidden":false},{"_id":"68886bb5af872d625c10c6d0","name":"Minda Hu","hidden":false},{"_id":"68886bb5af872d625c10c6d1","user":{"_id":"688969c4d0ac71994caa0df6","avatarUrl":"/avatars/f8f89e0ca810908cfe21c921d8373e28.svg","isPro":false,"fullname":"Huazheng Wang","user":"huazhengwang","type":"user"},"name":"Huazheng Wang","status":"claimed_verified","statusLastChangedAt":"2025-07-30T09:05:13.677Z","hidden":false},{"_id":"68886bb5af872d625c10c6d2","user":{"_id":"647fa485d0cd8be13e662973","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/647fa485d0cd8be13e662973/_QLgqLWhWir8_bkDCdTx6.png","isPro":false,"fullname":"Qingyun Wu","user":"qingyun-wu","type":"user"},"name":"Qingyun Wu","status":"claimed_verified","statusLastChangedAt":"2025-07-30T09:05:11.919Z","hidden":false},{"_id":"68886bb5af872d625c10c6d3","name":"Heng Ji","hidden":false},{"_id":"68886bb5af872d625c10c6d4","name":"Mengdi Wang","hidden":false}],"publishedAt":"2025-07-28T17:59:05.000Z","submittedOnDailyAt":"2025-07-29T05:06:58.475Z","title":"A Survey of Self-Evolving Agents: On Path to Artificial Super\n Intelligence","submittedOnDailyBy":{"_id":"65271df5abd7795aaa1fd86e","avatarUrl":"/avatars/99e1aabaefcf5e7f438817375cd1ddef.svg","isPro":false,"fullname":"Jiahao Qiu","user":"jiahaoq","type":"user"},"summary":"Large Language Models (LLMs) have demonstrated strong capabilities but remain\nfundamentally static, unable to adapt their internal parameters to novel tasks,\nevolving knowledge domains, or dynamic interaction contexts. As LLMs are\nincreasingly deployed in open-ended, interactive environments, this static\nnature has become a critical bottleneck, necessitating agents that can\nadaptively reason, act, and evolve in real time. This paradigm shift -- from\nscaling static models to developing self-evolving agents -- has sparked growing\ninterest in architectures and methods enabling continual learning and\nadaptation from data, interactions, and experiences. This survey provides the\nfirst systematic and comprehensive review of self-evolving agents, organized\naround three foundational dimensions -- what to evolve, when to evolve, and how\nto evolve. We examine evolutionary mechanisms across agent components (e.g.,\nmodels, memory, tools, architecture), categorize adaptation methods by stages\n(e.g., intra-test-time, inter-test-time), and analyze the algorithmic and\narchitectural designs that guide evolutionary adaptation (e.g., scalar rewards,\ntextual feedback, single-agent and multi-agent systems). Additionally, we\nanalyze evaluation metrics and benchmarks tailored for self-evolving agents,\nhighlight applications in domains such as coding, education, and healthcare,\nand identify critical challenges and research directions in safety,\nscalability, and co-evolutionary dynamics. By providing a structured framework\nfor understanding and designing self-evolving agents, this survey establishes a\nroadmap for advancing adaptive agentic systems in both research and real-world\ndeployments, ultimately shedding lights to pave the way for the realization of\nArtificial Super Intelligence (ASI), where agents evolve autonomously,\nperforming at or beyond human-level intelligence across a wide array of tasks.","upvotes":84,"discussionId":"68886bb6af872d625c10c6d5","githubRepo":"https://github.com/CharlesQ9/Self-Evolving-Agents","githubRepoAddedBy":"user","ai_summary":"This survey reviews self-evolving agents, focusing on mechanisms for adaptation in components like models, memory, and architecture, and discusses challenges in safety, scalability, and co-evolutionary dynamics.","ai_keywords":["self-evolving agents","continual learning","adaptation","intra-test-time","inter-test-time","scalar rewards","textual feedback","single-agent systems","multi-agent systems","evaluation metrics","benchmarks","coding","education","healthcare","Artificial Super Intelligence (ASI)"],"githubStars":902},"canReadDatabase":false,"canManagePapers":false,"canSubmit":false,"hasHfLevelAccess":false,"upvoted":false,"upvoters":[{"_id":"65271df5abd7795aaa1fd86e","avatarUrl":"/avatars/99e1aabaefcf5e7f438817375cd1ddef.svg","isPro":false,"fullname":"Jiahao Qiu","user":"jiahaoq","type":"user"},{"_id":"65f906e5c3dbdcae83ff7aac","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/65f906e5c3dbdcae83ff7aac/mdjiVkLDJgJcGLwv0rMe4.jpeg","isPro":false,"fullname":"Hongru Wang","user":"Merlin-Hongru","type":"user"},{"_id":"65f40e83653c231cbaf7defe","avatarUrl":"/avatars/afa5ce72324112739e539865c9aee26b.svg","isPro":false,"fullname":"Jiayi Zhang","user":"didiforhugface","type":"user"},{"_id":"6632160088f75d987d1a156f","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/6632160088f75d987d1a156f/mYlMQfK1BGWeEbOSMmeSb.jpeg","isPro":false,"fullname":"Hongzhang Liu","user":"Alphamasterliu","type":"user"},{"_id":"649ea7106282cb41e77760bc","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/649ea7106282cb41e77760bc/HlWjaqxr03ob93vdKg_LQ.jpeg","isPro":false,"fullname":"Isaac","user":"XiangJinYu","type":"user"},{"_id":"64b1303bf460afaefcf922c2","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/noauth/0zhdXXravx0CKJLTSVDwU.jpeg","isPro":false,"fullname":"Huan-ang Gao","user":"c7w","type":"user"},{"_id":"666aa99cd1652853e4f9a8b9","avatarUrl":"/avatars/7cd5a0c34b5ccb8eff5a353d88d15a93.svg","isPro":false,"fullname":"HanXiao","user":"HanXiao1999","type":"user"},{"_id":"6465d3bd63e7e09dd02e95c3","avatarUrl":"/avatars/b2798bd5f8368f956bf7fab79d9432f0.svg","isPro":false,"fullname":"Jie Feng","user":"JJ-TMT","type":"user"},{"_id":"620783f24e28382272337ba4","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/620783f24e28382272337ba4/zkUveQPNiDfYjgGhuFErj.jpeg","isPro":false,"fullname":"GuoLiangTang","user":"Tommy930","type":"user"},{"_id":"647eb24118274bce0308b2b8","avatarUrl":"/avatars/463e49a89b61164ccfad85ced10658b2.svg","isPro":false,"fullname":"Yiran Wu","user":"kevinwyr","type":"user"},{"_id":"66e2624a436a1798365e4581","avatarUrl":"/avatars/6c605807d34faa8fb505e135a4b47776.svg","isPro":false,"fullname":"Qihan Ren","user":"jasonrqh","type":"user"},{"_id":"641a7a8ff9dd6391a2457bbe","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/641a7a8ff9dd6391a2457bbe/_qdCnlrkpTfqZjXY4DFq_.jpeg","isPro":false,"fullname":"Zeyu Qin","user":"qqqzzzyyy","type":"user"}],"acceptLanguages":["*"],"dailyPaperRank":2}">

Papers

arxiv:2507.21046

A Survey of Self-Evolving Agents: On Path to Artificial Super Intelligence

Published on Jul 28, 2025

· Submitted by

Jiahao Qiu on Jul 29, 2025

#2 Paper of the day

Upvote

Authors:

Huan-ang Gao ,

Wenyue Hua ,

Hongzhang Liu ,

Han Xiao ,

Jiayi Zhang ,

Jinyu Xiang ,

Dongrui Liu ,

Qihan Ren ,

Cheng Qian ,

Abstract

This survey reviews self-evolving agents, focusing on mechanisms for adaptation in components like models, memory, and architecture, and discusses challenges in safety, scalability, and co-evolutionary dynamics.

AI-generated summary

Large Language Models (LLMs) have demonstrated strong capabilities but remain fundamentally static, unable to adapt their internal parameters to novel tasks, evolving knowledge domains, or dynamic interaction contexts. As LLMs are increasingly deployed in open-ended, interactive environments, this static nature has become a critical bottleneck, necessitating agents that can adaptively reason, act, and evolve in real time. This paradigm shift -- from scaling static models to developing self-evolving agents -- has sparked growing interest in architectures and methods enabling continual learning and adaptation from data, interactions, and experiences. This survey provides the first systematic and comprehensive review of self-evolving agents, organized around three foundational dimensions -- what to evolve, when to evolve, and how to evolve. We examine evolutionary mechanisms across agent components (e.g., models, memory, tools, architecture), categorize adaptation methods by stages (e.g., intra-test-time, inter-test-time), and analyze the algorithmic and architectural designs that guide evolutionary adaptation (e.g., scalar rewards, textual feedback, single-agent and multi-agent systems). Additionally, we analyze evaluation metrics and benchmarks tailored for self-evolving agents, highlight applications in domains such as coding, education, and healthcare, and identify critical challenges and research directions in safety, scalability, and co-evolutionary dynamics. By providing a structured framework for understanding and designing self-evolving agents, this survey establishes a roadmap for advancing adaptive agentic systems in both research and real-world deployments, ultimately shedding lights to pave the way for the realization of Artificial Super Intelligence (ASI), where agents evolve autonomously, performing at or beyond human-level intelligence across a wide array of tasks.