The authors introduce a 'Sleep' paradigm for LLMs, enabling continual learning through memory consolidation via knowledge seeding and a self-improvement 'Dreaming' process driven by reinforcement learning.</p>\n","updatedAt":"2026-06-03T02:09:10.018Z","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":0.869809627532959},"editors":["taesiri"],"editorAvatarUrls":["https://cdn-avatars.huggingface.co/v1/production/uploads/6039478ab3ecf716b1a5fd4d/_Thy4E7taiSYBLKxEKJbT.jpeg"],"reactions":[],"isReport":false}},{"id":"6a1fae1da2f8dbd3858552e3","author":{"_id":"6522c8cbdfc9ea4f31ca6fc2","avatarUrl":"/avatars/809a3d2e1779a02eea5f50d2a65e9544.svg","fullname":"Lane Sun","name":"lanesun","type":"user","isPro":false,"isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":2,"isUserFollowing":false},"createdAt":"2026-06-03T04:31:25.000Z","type":"comment","data":{"edited":false,"hidden":false,"latest":{"raw":"I feel that rather than being called \"sleep,\" this is more like a reflection mechanism. After all, this method does not seem to involve some kind of random connection in the latent space to achieve a creative emergence mechanism similar to human sleep. Please correct me if my understanding is wrong.","html":"<p>I feel that rather than being called \"sleep,\" this is more like a reflection mechanism. After all, this method does not seem to involve some kind of random connection in the latent space to achieve a creative emergence mechanism similar to human sleep. Please correct me if my understanding is wrong.</p>\n","updatedAt":"2026-06-03T04:31:25.741Z","author":{"_id":"6522c8cbdfc9ea4f31ca6fc2","avatarUrl":"/avatars/809a3d2e1779a02eea5f50d2a65e9544.svg","fullname":"Lane Sun","name":"lanesun","type":"user","isPro":false,"isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":2,"isUserFollowing":false}},"numEdits":0,"identifiedLanguage":{"language":"en","probability":0.9748892784118652},"editors":["lanesun"],"editorAvatarUrls":["/avatars/809a3d2e1779a02eea5f50d2a65e9544.svg"],"reactions":[],"isReport":false}}],"primaryEmailConfirmed":false,"paper":{"id":"2606.03979","authors":[{"_id":"6a1f8cb1e292c1c78ecb12f3","name":"Ali Behrouz","hidden":false},{"_id":"6a1f8cb1e292c1c78ecb12f4","name":"Farnoosh Hashemi","hidden":false},{"_id":"6a1f8cb1e292c1c78ecb12f5","name":"Vahab Mirrokni","hidden":false}],"publishedAt":"2026-06-02T00:00:00.000Z","submittedOnDailyAt":"2026-06-03T00:00:00.000Z","title":"Language Models Need Sleep: Learning to Self-Modify and Consolidate Memories","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":"The past few decades have witnessed significant advances in the design of machine learning algorithms, from early studies on task-specific shallow models to more general deep Large Language Models (LLMs). Despite showing promising results in tasks that require instant prediction or in-context learning, existing models lack the ability to continually learn and effectively transfer their temporal in-context knowledge to their long-term parameters. Inspired by human learning process, we introduce a ''Sleep'' paradigm that allows the models to continually learn, distill their short-term fragile memories into stable long-term knowledge with replay, and recursively improve themselves with ''Dreaming'' process. In more detail, sleep consists of two stages: (1) Memory Consolidation: an upward distillation process, called Knowledge Seeding, where the memories of a smaller-self are distilled into a larger network to provide more capacity while preserving the knowledge. As a proof of concept, we present a new Generalized Distillation process for {Knowledge Seeding} (i.e., the combination of on-policy distillation with Reinforcement Learning (RL)-based imitation learning); (2) Dreaming: a self-improvement phase, where the model uses RL to generate a curriculum of synthetic data to rehearse new knowledge and refine existing capabilities without human supervision. Our experiments on long-horizon, continual learning, knowledge incorporation, and few-shot generalization tasks support the importance of the sleep stage.","upvotes":9,"discussionId":"6a1f8cb1e292c1c78ecb12f6","ai_summary":"Deep learning models with sleep and dreaming paradigms enable continual learning through memory consolidation and self-improvement phases.","ai_keywords":["Large Language Models","continual learning","knowledge distillation","reinforcement learning","memory consolidation","dream phase","knowledge seeding","generalized distillation","synthetic data generation"],"ai_summary_model":"Qwen/Qwen2.5-Coder-32B-Instruct","organization":{"_id":"5e6aca39878b8b2bf9806447","name":"google","fullname":"Google","avatar":"https://cdn-avatars.huggingface.co/v1/production/uploads/5dd96eb166059660ed1ee413/WtA3YYitedOr9n02eHfJe.png"}},"canReadDatabase":false,"canManagePapers":false,"canSubmit":false,"hasHfLevelAccess":false,"upvoted":false,"upvoters":[{"_id":"6039478ab3ecf716b1a5fd4d","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/6039478ab3ecf716b1a5fd4d/_Thy4E7taiSYBLKxEKJbT.jpeg","isPro":true,"fullname":"taesiri","user":"taesiri","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":"63ca8e060609f1def7e6548a","avatarUrl":"/avatars/1da7947840cb87d5f77c0af9ee11f9c2.svg","isPro":true,"fullname":"Yi Jung","user":"YJ-142150","type":"user"},{"_id":"63180254212fce5a3cdc57a5","avatarUrl":"/avatars/9229d1ce9500f9b1a1ff1c4f6856ac10.svg","isPro":false,"fullname":"L","user":"TaidanaHito","type":"user"},{"_id":"6522c8cbdfc9ea4f31ca6fc2","avatarUrl":"/avatars/809a3d2e1779a02eea5f50d2a65e9544.svg","isPro":false,"fullname":"Lane Sun","user":"lanesun","type":"user"},{"_id":"64913f1b24d9bc9bb8ff407e","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/64913f1b24d9bc9bb8ff407e/N1cdMd9_DJb5GymdKJ3Mb.jpeg","isPro":false,"fullname":"Haoxiang Zhang","user":"IPF","type":"user"},{"_id":"63c1699e40a26dd2db32400d","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/63c1699e40a26dd2db32400d/3N0-Zp8igv8-52mXAdiiq.jpeg","isPro":false,"fullname":"Chroma","user":"Chroma111","type":"user"},{"_id":"65c20ee58aedd6edd2b89000","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/65c20ee58aedd6edd2b89000/LtS4YTbmxiCFqHSGHfdC8.png","isPro":false,"fullname":"Chmielewski","user":"Eryk-Chmielewski","type":"user"},{"_id":"64b0a5037a475fba70a7260d","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/64b0a5037a475fba70a7260d/MauBbb6raMA23yrR1Zq21.jpeg","isPro":false,"fullname":"Zhen Fang","user":"CostaliyA","type":"user"}],"acceptLanguages":["en"],"dailyPaperRank":0,"organization":{"_id":"5e6aca39878b8b2bf9806447","name":"google","fullname":"Google","avatar":"https://cdn-avatars.huggingface.co/v1/production/uploads/5dd96eb166059660ed1ee413/WtA3YYitedOr9n02eHfJe.png"},"markdownContentUrl":"https://huggingface.co/buckets/huggingchat/papers-content/resolve/2606/2606.03979.md"}">
Language Models Need Sleep: Learning to Self-Modify and Consolidate Memories
Abstract
Deep learning models with sleep and dreaming paradigms enable continual learning through memory consolidation and self-improvement phases.
The past few decades have witnessed significant advances in the design of machine learning algorithms, from early studies on task-specific shallow models to more general deep Large Language Models (LLMs). Despite showing promising results in tasks that require instant prediction or in-context learning, existing models lack the ability to continually learn and effectively transfer their temporal in-context knowledge to their long-term parameters. Inspired by human learning process, we introduce a ''Sleep'' paradigm that allows the models to continually learn, distill their short-term fragile memories into stable long-term knowledge with replay, and recursively improve themselves with ''Dreaming'' process. In more detail, sleep consists of two stages: (1) Memory Consolidation: an upward distillation process, called Knowledge Seeding, where the memories of a smaller-self are distilled into a larger network to provide more capacity while preserving the knowledge. As a proof of concept, we present a new Generalized Distillation process for {Knowledge Seeding} (i.e., the combination of on-policy distillation with Reinforcement Learning (RL)-based imitation learning); (2) Dreaming: a self-improvement phase, where the model uses RL to generate a curriculum of synthetic data to rehearse new knowledge and refine existing capabilities without human supervision. Our experiments on long-horizon, continual learning, knowledge incorporation, and few-shot generalization tasks support the importance of the sleep stage.
Community
The authors introduce a 'Sleep' paradigm for LLMs, enabling continual learning through memory consolidation via knowledge seeding and a self-improvement 'Dreaming' process driven by reinforcement learning.
I feel that rather than being called "sleep," this is more like a reflection mechanism. After all, this method does not seem to involve some kind of random connection in the latent space to achieve a creative emergence mechanism similar to human sleep. Please correct me if my understanding is wrong.
Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.
Tap or paste here to upload images
Cite arxiv.org/abs/2606.03979 in a model README.md to link it from this page.
Cite arxiv.org/abs/2606.03979 in a dataset README.md to link it from this page.
Cite arxiv.org/abs/2606.03979 in a Space README.md 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.