Librarian Bot. I found the following papers similar to this paper.

The following papers were recommended by the Semantic Scholar API

PRISM: Purified Representation and Integrated Semantic Modeling for Generative Sequential Recommendation (2026)
UniGRec: Unified Generative Recommendation with Soft Identifiers for End-to-End Optimization (2026)
Masked Diffusion Generative Recommendation (2026)
Multimodal Generative Recommendation for Fusing Semantic and Collaborative Signals (2026)
HiGR: Efficient Generative Slate Recommendation via Hierarchical Planning and Multi-Objective Preference Alignment (2025)
Unleashing the Native Recommendation Potential: LLM-Based Generative Recommendation via Structured Term Identifiers (2026)
CEMG: Collaborative-Enhanced Multimodal Generative Recommendation (2025)

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":"2026-02-05T01:43:53.581Z","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":0.7166656255722046},"editors":["librarian-bot"],"editorAvatarUrls":["https://cdn-avatars.huggingface.co/v1/production/uploads/1674830754237-63d3e0e8ff1384ce6c5dd17d.jpeg"],"reactions":[],"isReport":false}}],"primaryEmailConfirmed":false,"paper":{"id":"2602.02338","authors":[{"_id":"698228299c2f139721ec350e","user":{"_id":"694e1505a624c194a198a20c","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/694e1505a624c194a198a20c/ZS98Y8FcGS3CQOH2wnqkd.jpeg","isPro":false,"fullname":"Yu Liang","user":"Ricardoly","type":"user"},"name":"Yu Liang","status":"claimed_verified","statusLastChangedAt":"2026-02-04T19:03:59.622Z","hidden":false},{"_id":"698228299c2f139721ec350f","name":"Zhongjin Zhang","hidden":false},{"_id":"698228299c2f139721ec3510","user":{"_id":"641f18a673cfc036ddbaeccf","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/no-auth/RtOK_qV71NM87A67XFJCN.png","isPro":false,"fullname":"Yuxuan Zhu","user":"Ethan7","type":"user"},"name":"Yuxuan Zhu","status":"claimed_verified","statusLastChangedAt":"2026-02-04T12:31:16.557Z","hidden":false},{"_id":"698228299c2f139721ec3511","name":"Kerui Zhang","hidden":false},{"_id":"698228299c2f139721ec3512","name":"Zhiluohan Guo","hidden":false},{"_id":"698228299c2f139721ec3513","name":"Wenhang Zhou","hidden":false},{"_id":"698228299c2f139721ec3514","name":"Zonqi Yang","hidden":false},{"_id":"698228299c2f139721ec3515","name":"Kangle Wu","hidden":false},{"_id":"698228299c2f139721ec3516","name":"Yabo Ni","hidden":false},{"_id":"698228299c2f139721ec3517","name":"Anxiang Zeng","hidden":false},{"_id":"698228299c2f139721ec3518","name":"Cong Fu","hidden":false},{"_id":"698228299c2f139721ec3519","name":"Jianxin Wang","hidden":false},{"_id":"698228299c2f139721ec351a","name":"Jiazhi Xia","hidden":false}],"publishedAt":"2026-02-02T17:00:04.000Z","submittedOnDailyAt":"2026-02-03T14:30:54.265Z","title":"Rethinking Generative Recommender Tokenizer: Recsys-Native Encoding and Semantic Quantization Beyond LLMs","submittedOnDailyBy":{"_id":"694e1505a624c194a198a20c","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/694e1505a624c194a198a20c/ZS98Y8FcGS3CQOH2wnqkd.jpeg","isPro":false,"fullname":"Yu Liang","user":"Ricardoly","type":"user"},"summary":"Semantic ID (SID)-based recommendation is a promising paradigm for scaling sequential recommender systems, but existing methods largely follow a semantic-centric pipeline: item embeddings are learned from foundation models and discretized using generic quantization schemes. This design is misaligned with generative recommendation objectives: semantic embeddings are weakly coupled with collaborative prediction, and generic quantization is inefficient at reducing sequential uncertainty for autoregressive modeling. To address these, we propose ReSID, a recommendation-native, principled SID framework that rethinks representation learning and quantization from the perspective of information preservation and sequential predictability, without relying on LLMs. ReSID consists of two components: (i) Field-Aware Masked Auto-Encoding (FAMAE), which learns predictive-sufficient item representations from structured features, and (ii) Globally Aligned Orthogonal Quantization (GAOQ), which produces compact and predictable SID sequences by jointly reducing semantic ambiguity and prefix-conditional uncertainty. Theoretical analysis and extensive experiments across ten datasets show the effectiveness of ReSID. ReSID consistently outperforms strong sequential and SID-based generative baselines by an average of over 10%, while reducing tokenization cost by up to 122x. Code is available at https://github.com/FuCongResearchSquad/ReSID.","upvotes":40,"discussionId":"698228299c2f139721ec351b","githubRepo":"https://github.com/FuCongResearchSquad/ReSID","githubRepoAddedBy":"user","ai_summary":"ReSID presents a novel recommendation-native framework that improves sequential recommendation by learning predictive item representations and optimizing quantization for information preservation and sequential predictability.","ai_keywords":["semantic ID","sequential recommender systems","generative recommendation","foundation models","quantization","masked auto-encoding","orthogonal quantization","information preservation","sequential predictability","collaborative prediction","autoregressive modeling"],"githubStars":4,"organization":{"_id":"693bcc3aca53da32fcf95a61","name":"PIIR","fullname":"Personalized&Inclusive Intelligence Research"}},"canReadDatabase":false,"canManagePapers":false,"canSubmit":false,"hasHfLevelAccess":false,"upvoted":false,"upvoters":[{"_id":"68d210738c6103d703e915a8","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/no-auth/9kAT_FpLNTKN3uld5yMl5.png","isPro":false,"fullname":"zhaoyu","user":"SmallYelloHair","type":"user"},{"_id":"6982a47e3ea114b8ee075403","avatarUrl":"/avatars/c9fecc0ea7f8e1a55b62922d90b25f42.svg","isPro":false,"fullname":"Hao Zhipeng","user":"hzp930930","type":"user"},{"_id":"666a55b1ae0b4ceca4e8ff86","avatarUrl":"/avatars/642e703469c8c2d91f6011879747f7dd.svg","isPro":false,"fullname":"gawa","user":"gawa123","type":"user"},{"_id":"68d20f967ded6497d233f808","avatarUrl":"/avatars/88505147ab02669aeed47a0103d45330.svg","isPro":false,"fullname":"yzq","user":"Inconnu9","type":"user"},{"_id":"68d214fe2e59f503dbffb8c9","avatarUrl":"/avatars/fef3c37dc9dd9e4833d7ba05c9e293de.svg","isPro":false,"fullname":"guo","user":"guoehan","type":"user"},{"_id":"67e21ce15d02fcdf8d7abd39","avatarUrl":"/avatars/01f63e591d327f3d54f51460a599dc80.svg","isPro":false,"fullname":"kun","user":"vicowang","type":"user"},{"_id":"69427905aca980710cb83e9d","avatarUrl":"/avatars/176a9ff6cc1abfd056e4fee206136ea4.svg","isPro":false,"fullname":"catfish","user":"catfish2357","type":"user"},{"_id":"68d218c230c41a74dbee3564","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/no-auth/SjHI-qhzku51Sod3AUKxm.png","isPro":false,"fullname":"Jingzhi.Ding","user":"Manderin","type":"user"},{"_id":"677fdca602e79f5b6b2cbc18","avatarUrl":"/avatars/223ce20e47fc70cbe80b2718a0017fec.svg","isPro":false,"fullname":"ShiYifan","user":"lancelot910520","type":"user"},{"_id":"641f18a673cfc036ddbaeccf","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/no-auth/RtOK_qV71NM87A67XFJCN.png","isPro":false,"fullname":"Yuxuan Zhu","user":"Ethan7","type":"user"},{"_id":"696f1f0d9e870f6dc6462932","avatarUrl":"/avatars/dfdd0ee6736ed8cb9f3397cc5459e4f4.svg","isPro":false,"fullname":"CaptainKarl","user":"CaptainKarl","type":"user"},{"_id":"6411be516a43f40dec572448","avatarUrl":"/avatars/9530f44031ac6d76cc5fb87cf37f281e.svg","isPro":false,"fullname":"MingguangShao","user":"shao1995","type":"user"}],"acceptLanguages":["*"],"dailyPaperRank":0,"organization":{"_id":"693bcc3aca53da32fcf95a61","name":"PIIR","fullname":"Personalized&Inclusive Intelligence Research"}}">

Papers

arxiv:2602.02338

Rethinking Generative Recommender Tokenizer: Recsys-Native Encoding and Semantic Quantization Beyond LLMs

Published on Feb 2

· Submitted by

Yu Liang on Feb 3

· Personalized&Inclusive Intelligence Research

Upvote

Authors:

Yu Liang ,

Yuxuan Zhu ,

Abstract

ReSID presents a novel recommendation-native framework that improves sequential recommendation by learning predictive item representations and optimizing quantization for information preservation and sequential predictability.

AI-generated summary

Semantic ID (SID)-based recommendation is a promising paradigm for scaling sequential recommender systems, but existing methods largely follow a semantic-centric pipeline: item embeddings are learned from foundation models and discretized using generic quantization schemes. This design is misaligned with generative recommendation objectives: semantic embeddings are weakly coupled with collaborative prediction, and generic quantization is inefficient at reducing sequential uncertainty for autoregressive modeling. To address these, we propose ReSID, a recommendation-native, principled SID framework that rethinks representation learning and quantization from the perspective of information preservation and sequential predictability, without relying on LLMs. ReSID consists of two components: (i) Field-Aware Masked Auto-Encoding (FAMAE), which learns predictive-sufficient item representations from structured features, and (ii) Globally Aligned Orthogonal Quantization (GAOQ), which produces compact and predictable SID sequences by jointly reducing semantic ambiguity and prefix-conditional uncertainty. Theoretical analysis and extensive experiments across ten datasets show the effectiveness of ReSID. ReSID consistently outperforms strong sequential and SID-based generative baselines by an average of over 10%, while reducing tokenization cost by up to 122x. Code is available at https://github.com/FuCongResearchSquad/ReSID.