https://github.com/Dereck0602/Awesome_Test_Time_LLMs

\n","updatedAt":"2025-01-07T04:05:27.993Z","author":{"_id":"60f1abe7544c2adfd699860c","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/1674929746905-60f1abe7544c2adfd699860c.jpeg","fullname":"AK","name":"akhaliq","type":"user","isPro":false,"isHf":true,"isHfAdmin":false,"isMod":false,"followerCount":9179,"isUserFollowing":false}},"numEdits":0,"identifiedLanguage":{"language":"en","probability":0.8219125866889954},"editors":["akhaliq"],"editorAvatarUrls":["https://cdn-avatars.huggingface.co/v1/production/uploads/1674929746905-60f1abe7544c2adfd699860c.jpeg"],"reactions":[],"isReport":false}},{"id":"677dd5e03b6523609213c696","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":"2025-01-08T01:33:20.000Z","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* [Can OpenAI o1 outperform humans in higher-order cognitive thinking?](https://huggingface.co/papers/2412.05753) (2024)\n* [Generative AI as a Tool or Leader? Exploring AI-Augmented Thinking in Student Programming Tasks](https://huggingface.co/papers/2411.19490) (2024)\n* [The CTSkills App - Measuring Problem Decomposition Skills of Students in Computational Thinking](https://huggingface.co/papers/2411.14945) (2024)\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":"2025-01-08T01:33:20.398Z","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":0.8056175708770752},"editors":["librarian-bot"],"editorAvatarUrls":["https://cdn-avatars.huggingface.co/v1/production/uploads/1674830754237-63d3e0e8ff1384ce6c5dd17d.jpeg"],"reactions":[],"isReport":false}}],"primaryEmailConfirmed":false,"paper":{"id":"2501.02497","authors":[{"_id":"677ca7db0c9718b04a426305","user":{"_id":"63412a43a7582111c3f1cadd","avatarUrl":"/avatars/033e809a7aa3d9c4b1326e9195290f65.svg","isPro":false,"fullname":"Yixin Ji","user":"Yisam","type":"user"},"name":"Yixin Ji","status":"extracted_confirmed","statusLastChangedAt":"2025-01-07T04:53:11.414Z","hidden":false},{"_id":"677ca7db0c9718b04a426306","user":{"_id":"6670e285b0c03c4e9d6e0985","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/6670e285b0c03c4e9d6e0985/j9Zr-lOtrRmpFz5f4x420.jpeg","isPro":false,"fullname":"Juntao Li","user":"douvleplus","type":"user"},"name":"Juntao Li","status":"admin_assigned","statusLastChangedAt":"2025-01-07T09:51:35.155Z","hidden":false},{"_id":"677ca7db0c9718b04a426307","user":{"_id":"6239888e7fef05b7bdd5fcff","avatarUrl":"/avatars/54fcc756b8c0936b6bb410c6e0e02d75.svg","isPro":false,"fullname":"Hai Ye","user":"oceanpty","type":"user"},"name":"Hai Ye","status":"claimed_verified","statusLastChangedAt":"2025-01-07T13:02:02.646Z","hidden":false},{"_id":"677ca7db0c9718b04a426308","user":{"_id":"677f1a2926d2ab07cfc7b37e","avatarUrl":"/avatars/fca140098b3aa41b93107603b941f35f.svg","isPro":false,"fullname":"Wu Kaixin","user":"daniel-wkx","type":"user"},"name":"Kaixin Wu","status":"claimed_verified","statusLastChangedAt":"2025-01-09T10:07:22.984Z","hidden":false},{"_id":"677ca7db0c9718b04a426309","name":"Jia Xu","hidden":false},{"_id":"677ca7db0c9718b04a42630a","name":"Linjian Mo","hidden":false},{"_id":"677ca7db0c9718b04a42630b","name":"Min Zhang","hidden":false}],"publishedAt":"2025-01-05T10:24:20.000Z","submittedOnDailyAt":"2025-01-07T01:35:27.984Z","title":"Test-time Computing: from System-1 Thinking to System-2 Thinking","submittedOnDailyBy":{"_id":"60f1abe7544c2adfd699860c","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/1674929746905-60f1abe7544c2adfd699860c.jpeg","isPro":false,"fullname":"AK","user":"akhaliq","type":"user"},"summary":"The remarkable performance of the o1 model in complex reasoning demonstrates\nthat test-time computing scaling can further unlock the model's potential,\nenabling powerful System-2 thinking. However, there is still a lack of\ncomprehensive surveys for test-time computing scaling. We trace the concept of\ntest-time computing back to System-1 models. In System-1 models, test-time\ncomputing addresses distribution shifts and improves robustness and\ngeneralization through parameter updating, input modification, representation\nediting, and output calibration. In System-2 models, it enhances the model's\nreasoning ability to solve complex problems through repeated sampling,\nself-correction, and tree search. We organize this survey according to the\ntrend of System-1 to System-2 thinking, highlighting the key role of test-time\ncomputing in the transition from System-1 models to weak System-2 models, and\nthen to strong System-2 models. We also point out a few possible future\ndirections.","upvotes":45,"discussionId":"677ca7dc0c9718b04a426342","githubRepo":"https://github.com/dereck0602/awesome_test_time_llms","githubRepoAddedBy":"auto","ai_summary":"Test-time computing scaling enhances complex reasoning in models, improving robustness, generalization, and problem-solving from System-1 to System-2 thinking through various mechanisms.","ai_keywords":["test-time computing","System-1 models","distribution shifts","robustness","generalization","parameter updating","input modification","representation editing","output calibration","System-2 models","repeated sampling","self-correction","tree search"],"githubStars":144},"canReadDatabase":false,"canManagePapers":false,"canSubmit":false,"hasHfLevelAccess":false,"upvoted":false,"upvoters":[{"_id":"65ddc21ffd522989011dd0b6","avatarUrl":"/avatars/31101d2ce37f00c4f8f1f43f2c793dfb.svg","isPro":false,"fullname":"KyleShi","user":"KyleShi","type":"user"},{"_id":"63082bb7bc0a2a5ee2253523","avatarUrl":"/avatars/6cf8d12d16d15db1070fbea89b5b3967.svg","isPro":false,"fullname":"Kuo-Hsin Tu","user":"dapumptu","type":"user"},{"_id":"626cf0f65651e31a7a2b9779","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/626cf0f65651e31a7a2b9779/xES0qcewzvfSIy_0Sqd0u.png","isPro":false,"fullname":"Ding","user":"dyyyyyyyy","type":"user"},{"_id":"655d9ff875956386715ccaae","avatarUrl":"/avatars/75f7bef3be84dd40d11254757b15bfd3.svg","isPro":false,"fullname":"Hu","user":"pbhu1024","type":"user"},{"_id":"63b6f2e752c02ae8acbaa4d8","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/1672934038280-noauth.jpeg","isPro":false,"fullname":"Habibullah Akbar","user":"ChavyvAkvar","type":"user"},{"_id":"6486bb3d4c025cf3c41e7767","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/6486bb3d4c025cf3c41e7767/1I6yYhv06rHX1_1iIOdjP.png","isPro":false,"fullname":"Quentin Tardif","user":"ntnq","type":"user"},{"_id":"60d1d7f94672e8c13e7c2bce","avatarUrl":"/avatars/be3a3990492ef1c5f3ff33cc043ae166.svg","isPro":false,"fullname":"Juntao Li","user":"ljtsuda","type":"user"},{"_id":"6239888e7fef05b7bdd5fcff","avatarUrl":"/avatars/54fcc756b8c0936b6bb410c6e0e02d75.svg","isPro":false,"fullname":"Hai Ye","user":"oceanpty","type":"user"},{"_id":"619550cb4f08530be9bec2cf","avatarUrl":"/avatars/d1df21b0ff7cfa098d6a0c5375e2c194.svg","isPro":false,"fullname":"Luong Son Tinh","user":"convoicon","type":"user"},{"_id":"648eb1eb59c4e5c87dc116e0","avatarUrl":"/avatars/c636cea39c2c0937f01398c94ead5dad.svg","isPro":false,"fullname":"fdsqefsgergd","user":"T-representer","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":"6101c620900eaa0057c2ce1d","avatarUrl":"/avatars/bd282166c120711c65b5409dc860ac58.svg","isPro":false,"fullname":"Abdel-Dayane Marcos","user":"admarcosai","type":"user"}],"acceptLanguages":["*"],"dailyPaperRank":2}">

Papers

arxiv:2501.02497

Test-time Computing: from System-1 Thinking to System-2 Thinking

Published on Jan 5, 2025

· Submitted by

AK on Jan 7, 2025

#2 Paper of the day

Upvote

Authors:

Yixin Ji ,

Juntao Li ,

Hai Ye ,

Kaixin Wu ,

Abstract

Test-time computing scaling enhances complex reasoning in models, improving robustness, generalization, and problem-solving from System-1 to System-2 thinking through various mechanisms.

AI-generated summary

The remarkable performance of the o1 model in complex reasoning demonstrates that test-time computing scaling can further unlock the model's potential, enabling powerful System-2 thinking. However, there is still a lack of comprehensive surveys for test-time computing scaling. We trace the concept of test-time computing back to System-1 models. In System-1 models, test-time computing addresses distribution shifts and improves robustness and generalization through parameter updating, input modification, representation editing, and output calibration. In System-2 models, it enhances the model's reasoning ability to solve complex problems through repeated sampling, self-correction, and tree search. We organize this survey according to the trend of System-1 to System-2 thinking, highlighting the key role of test-time computing in the transition from System-1 models to weak System-2 models, and then to strong System-2 models. We also point out a few possible future directions.