mlsanityData to 7 Oct 2026

Learnable Item Tokenization for Generative Recommendation

arXiv 2405.07314 ↗First posted 12 May 2024cs.IRcited by 106

Not on the map. Why?

Abstract (arXiv)

Utilizing powerful Large Language Models (LLMs) for generative recommendation has attracted much attention. Nevertheless, a crucial challenge is transforming recommendation data into the language space of LLMs through effective item tokenization. Current approaches, such as ID, textual, and codebook-based identifiers, exhibit shortcomings in encoding semantic information, incorporating collaborative signals, or handling code assignment bias. To address these limitations, we propose LETTER (a LEarnable Tokenizer for generaTivE Recommendation), which integrates hierarchical semantics, collaborative signals, and code assignment diversity to satisfy the essential requirements of identifiers. LETTER incorporates Residual Quantized VAE for semantic regularization, a contrastive alignment loss for collaborative regularization, and a diversity loss to mitigate code assignment bias. We instantiate LETTER on two models and propose a ranking-guided generation loss to augment their ranking ability theoretically. Experiments on three datasets validate the superiority of LETTER, advancing the state-of-the-art in the field of LLM-based generative recommendation.

Key facts. Cited by 106 papers in the map (data through 7 Oct 2026). First posted on arXiv 12 May 2024.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2024: 5; 2025: 43; 2026: 58.

106 papers.

Sort:

By area:

  1. 2025Sparse Meets Dense: Unified Generative Recommendations with Cascaded Sparse-Dense Representationscited by 35
  2. 2025ActionPiece: Contextually Tokenizing Action Sequences for Generative Recommendationcited by 18
  3. 2025R4ec: A Reasoning, Reflection, and Refinement Framework for Recommendation Systemscited by 16
  4. 2025Multimodal Medical Code Tokenizercited by 14
  5. 2025R2ec: Towards Large Recommender Models with Reasoningcited by 13
  6. 2025Slow Thinking for Sequential Recommendationcited by 11
  7. 2025HiD-VAE: Interpretable Generative Recommendation via Hierarchical and Disentangled Semantic IDscited by 11
  8. 2025MMQ: Multimodal Mixture-of-Quantization Tokenization for Semantic ID Generation and User Behavioral Adaptationcited by 10
  9. 2025CoFiRec: Coarse-to-Fine Tokenization for Generative Recommendationcited by 10
  10. 2024Enhancing Item Tokenization for Generative Recommendation through Self-Improvementcited by 9
  11. 2025LLaDA-Rec: Discrete Diffusion for Parallel Semantic ID Generation in Generative Recommendationcited by 9
  12. 2024Unleash LLMs Potential for Sequential Recommendation by Coordinating Dual Dynamic Index Mechanismcited by 8
38 more
  1. 2024Causality-Enhanced Behavior Sequence Modeling in LLMs for Personalized Recommendationcited by 8
  2. 2025Universal Item Tokenization for Transferable Generative Recommendationcited by 7
  3. 2025GRAM: Generative Recommendation via Semantic-aware Multi-granular Late Fusioncited by 7
  4. 2025Discrete Tokenization for Multimodal LLMs: A Comprehensive Surveycited by 7
  5. 2026Generative Recommendation for Large-Scale Advertisingcited by 7
  6. 2024Preference Discerning with LLM-Enhanced Generative Retrievalcited by 6
  7. 2025Eliminating Out-of-Domain Recommendations in LLM-based Recommender Systems: A Unified Viewcited by 5
  8. 2025OneLoc: Geo-Aware Generative Recommender Systems for Local Life Servicecited by 5
  9. 2025NEZHA: A Zero-sacrifice and Hyperspeed Decoding Architecture for Generative Recommendationscited by 5
  10. 2025Masked Diffusion for Generative Recommendationcited by 5
  11. 2026Stop Treating Collisions Equally: Qualification-Aware Semantic ID Learning for Recommendation at Industrial Scalecited by 5
  12. 2026GenRec: A Preference-Oriented Generative Framework for Large-Scale Recommendationcited by 5
  13. 2025Large Language Model as Universal Retriever in Industrial-Scale Recommender Systemcited by 4
  14. 2025Pre-training Generative Recommender with Multi-Identifier Item Tokenizationcited by 4
  15. 2025A Simple Contrastive Framework Of Item Tokenization For Generative Recommendationcited by 4
  16. 2026CRAB: Codebook Rebalancing for Bias Mitigation in Generative Recommendationcited by 4
  17. 2025Can LLMs Outshine Conventional Recommenders? A Comparative Evaluationcited by 3
  18. 2025Beyond Unimodal Boundaries: Generative Recommendation with Multimodal Semanticscited by 3
  19. 2025GFlowGR: Fine-tuning Generative Recommendation Frameworks with Generative Flow Networkscited by 3
  20. 2025Pctx: Tokenizing Personalized Context for Generative Recommendationcited by 3
  21. 2025Multi-Aspect Cross-modal Quantization for Generative Recommendationcited by 3
  22. 2026When Text-as-Vision Meets Semantic IDs in Generative Recommendation: An Empirical Studycited by 3
  23. 2026End-to-End Semantic ID Generation for Generative Advertisement Recommendationcited by 3
  24. 2026Understanding Semantic IDs: From Item Representation to Item Selection in Generative Recommendationcited by 3
  25. 2024MVIGER: Multi-View Variational Integration of Complementary Knowledge for Generative Recommendercited by 2
  26. 2025DGenCTR: Towards a Universal Generative Paradigm for Click-Through Rate Prediction via Discrete Diffusioncited by 2
  27. 2025Empowering Large Language Model for Sequential Recommendation via Multimodal Embeddings and Semantic IDscited by 2
  28. 2025Enhancing Time Awareness in Generative Recommendationcited by 2
  29. 2025From Past To Path: Masked History Learning for Next-Item Prediction in Generative Recommendationcited by 2
  30. 2025Cognitive-Aligned Spatio-Temporal Large Language Models For Next Point-of-Interest Predictioncited by 2
  31. 2025Continual Low-Rank Adapters for LLM-based Generative Recommender Systemscited by 2
  32. 2026Rethinking Generative Recommender Tokenizer: Recsys-Native Encoding and Semantic Quantization Beyond LLMscited by 2
  33. 2026Verifiable Reasoning for LLM-based Generative Recommendationcited by 2
  34. 2026Grounded Token Initialization for New Vocabulary in LMs for Generative Recommendationcited by 2
  35. 2026SAPO: Step-Aligned Policy Optimization for Reasoning-Based Generative Recommendationcited by 2
  36. 2026LLMs Need Encoders for Semantic IDs Toocited by 2
  37. 2026TRACER: Token ReAssignment for Concept ERasure in Generative Recommendationcited by 2
  38. 2026HoloRec: Holistic Encoding and Interleaved Reasoning for Generative Recommendationcited by 2

50 most cited of 106 All 106 (JSON)