Hugging Face Daily Papers · · 4 min read

MLEvolve: A Self-Evolving Framework for Automated Machine Learning Algorithm Discovery

Mirrored from Hugging Face Daily Papers for archival readability. Support the source by reading on the original site.

MLEvolve is an LLM-based multi-agent framework for automated machine learning algorithm discovery, featuring Progressive Monte Carlo Graph Search and retrospective memory to enhance long-horizon optimization performance.</p>\n","updatedAt":"2026-06-05T01:59:44.126Z","author":{"_id":"6039478ab3ecf716b1a5fd4d","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/6039478ab3ecf716b1a5fd4d/_Thy4E7taiSYBLKxEKJbT.jpeg","fullname":"taesiri","name":"taesiri","type":"user","isPro":true,"isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":310,"isUserFollowing":false}},"numEdits":0,"identifiedLanguage":{"language":"en","probability":0.8330307006835938},"editors":["taesiri"],"editorAvatarUrls":["https://cdn-avatars.huggingface.co/v1/production/uploads/6039478ab3ecf716b1a5fd4d/_Thy4E7taiSYBLKxEKJbT.jpeg"],"reactions":[],"isReport":false}}],"primaryEmailConfirmed":false,"paper":{"id":"2606.06473","authors":[{"_id":"6a222d6a3490a593e87b13fa","name":"Shangheng Du","hidden":false},{"_id":"6a222d6a3490a593e87b13fb","name":"Xiangchao Yan","hidden":false},{"_id":"6a222d6a3490a593e87b13fc","name":"Jinxin Shi","hidden":false},{"_id":"6a222d6a3490a593e87b13fd","name":"Zongsheng Cao","hidden":false},{"_id":"6a222d6a3490a593e87b13fe","name":"Shiyang Feng","hidden":false},{"_id":"6a222d6a3490a593e87b13ff","name":"Zichen Liang","hidden":false},{"_id":"6a222d6a3490a593e87b1400","name":"Boyuan Sun","hidden":false},{"_id":"6a222d6a3490a593e87b1401","name":"Tianshuo Peng","hidden":false},{"_id":"6a222d6a3490a593e87b1402","name":"Yifan Zhou","hidden":false},{"_id":"6a222d6a3490a593e87b1403","name":"Xin Li","hidden":false},{"_id":"6a222d6a3490a593e87b1404","name":"Jie Zhou","hidden":false},{"_id":"6a222d6a3490a593e87b1405","name":"Liang He","hidden":false},{"_id":"6a222d6a3490a593e87b1406","name":"Bo Zhang","hidden":false},{"_id":"6a222d6a3490a593e87b1407","name":"Lei Bai","hidden":false}],"publishedAt":"2026-06-04T00:00:00.000Z","submittedOnDailyAt":"2026-06-05T00:00:00.000Z","title":"MLEvolve: A Self-Evolving Framework for Automated Machine Learning Algorithm Discovery","submittedOnDailyBy":{"_id":"6039478ab3ecf716b1a5fd4d","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/6039478ab3ecf716b1a5fd4d/_Thy4E7taiSYBLKxEKJbT.jpeg","isPro":true,"fullname":"taesiri","user":"taesiri","type":"user","name":"taesiri"},"summary":"Large language model (LLM) agents are increasingly applied to long-horizon tasks such as scientific discovery and machine learning engineering (MLE), where sustained self-evolution becomes a key capability. However, existing MLE agents suffer from inter-branch information isolation, memoryless search, and lack of hierarchical control, which together hinder long-horizon optimization. We present MLEvolve, an LLM-based self-evolving multi-agent framework for end-to-end machine learning algorithm discovery. By extending tree search to Progressive MCGS, MLEvolve enables cross-branch information flow through graph-based reference edges and gradually shifts the search from broad exploration to focused exploitation with an entropy-inspired progressive schedule. To allow the agent to evolve with accumulated experience, we introduce Retrospective Memory, which combines a cold-start domain knowledge base with a dynamic global memory for task-specific experience retrieval and reuse. For stable long-horizon iteration, we further decouple strategic planning from code generation with adaptive coding modes. Evaluation on MLE-Bench shows that MLEvolve achieves state-of-the-art performance across multiple dimensions including average medal rate and valid submission rate under a 12-hour budget (half the standard runtime). Moreover, MLEvolve also outperforms specialized algorithm discovery methods including AlphaEvolve on mathematical algorithm optimization tasks, demonstrating strong cross-domain generalization. Our code is available at https://github.com/InternScience/MLEvolve.","upvotes":2,"discussionId":"6a222d6b3490a593e87b1408","githubRepo":"https://github.com/InternScience/MLEvolve","githubRepoAddedBy":"user","ai_summary":"MLEvolve is an LLM-based multi-agent framework that enables long-horizon machine learning algorithm discovery through improved search mechanisms, memory systems, and adaptive coding strategies.","ai_keywords":["large language model","multi-agent framework","machine learning algorithm discovery","tree search","Progressive MCGS","graph-based reference edges","entropy-inspired progressive schedule","Retrospective Memory","dynamic global memory","strategic planning","code generation","adaptive coding modes","MLE-Bench","AlphaEvolve"],"ai_summary_model":"Qwen/Qwen2.5-Coder-32B-Instruct","githubStars":296},"canReadDatabase":false,"canManagePapers":false,"canSubmit":false,"hasHfLevelAccess":false,"upvoted":false,"upvoters":[{"_id":"67cfd5f4d8cb8688d7e2df22","avatarUrl":"/avatars/099139aac6d803fa47579a1152da39ef.svg","isPro":false,"fullname":"Songtao Huang","user":"huangst","type":"user"},{"_id":"620783f24e28382272337ba4","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/620783f24e28382272337ba4/zkUveQPNiDfYjgGhuFErj.jpeg","isPro":false,"fullname":"GuoLiangTang","user":"Tommy930","type":"user"}],"acceptLanguages":["en"],"dailyPaperRank":0,"markdownContentUrl":"https://huggingface.co/buckets/huggingchat/papers-content/resolve/2606/2606.06473.md"}">
Papers
arxiv:2606.06473

MLEvolve: A Self-Evolving Framework for Automated Machine Learning Algorithm Discovery

Published on Jun 4
· Submitted by
taesiri
on Jun 5
Authors:
,
,
,
,
,
,
,
,
,
,
,
,
,

Abstract

MLEvolve is an LLM-based multi-agent framework that enables long-horizon machine learning algorithm discovery through improved search mechanisms, memory systems, and adaptive coding strategies.

Large language model (LLM) agents are increasingly applied to long-horizon tasks such as scientific discovery and machine learning engineering (MLE), where sustained self-evolution becomes a key capability. However, existing MLE agents suffer from inter-branch information isolation, memoryless search, and lack of hierarchical control, which together hinder long-horizon optimization. We present MLEvolve, an LLM-based self-evolving multi-agent framework for end-to-end machine learning algorithm discovery. By extending tree search to Progressive MCGS, MLEvolve enables cross-branch information flow through graph-based reference edges and gradually shifts the search from broad exploration to focused exploitation with an entropy-inspired progressive schedule. To allow the agent to evolve with accumulated experience, we introduce Retrospective Memory, which combines a cold-start domain knowledge base with a dynamic global memory for task-specific experience retrieval and reuse. For stable long-horizon iteration, we further decouple strategic planning from code generation with adaptive coding modes. Evaluation on MLE-Bench shows that MLEvolve achieves state-of-the-art performance across multiple dimensions including average medal rate and valid submission rate under a 12-hour budget (half the standard runtime). Moreover, MLEvolve also outperforms specialized algorithm discovery methods including AlphaEvolve on mathematical algorithm optimization tasks, demonstrating strong cross-domain generalization. Our code is available at https://github.com/InternScience/MLEvolve.

Community

Paper submitter about 9 hours ago

MLEvolve is an LLM-based multi-agent framework for automated machine learning algorithm discovery, featuring Progressive Monte Carlo Graph Search and retrospective memory to enhance long-horizon optimization performance.

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.
Tap or paste here to upload images

· Sign up or log in to comment

Get this paper in your agent:

hf papers read 2606.06473
Don't have the latest CLI?
curl -LsSf https://hf.co/cli/install.sh | bash

Models citing this paper 0

No model linking this paper

Cite arxiv.org/abs/2606.06473 in a model README.md to link it from this page.

Datasets citing this paper 0

No dataset linking this paper

Cite arxiv.org/abs/2606.06473 in a dataset README.md to link it from this page.

Spaces citing this paper 0

No Space linking this paper

Cite arxiv.org/abs/2606.06473 in a Space README.md to link it from this page.

Collections including this paper 0

No Collection including this paper

Add this paper to a collection to link it from this page.

Discussion (0)

Sign in to join the discussion. Free account, 30 seconds — email code or GitHub.

Sign in →

No comments yet. Sign in and be the first to say something.

More from Hugging Face Daily Papers