A curated, continuously updated collection of research on Large Language Models for Recommendation —
spanning Generative Recommendation and Agentic Recommendation
Last Updated:Loading... Auto-updated daily at 8:05 AM Beijing Time | Tracking latest research from ArXiv
Filter:
Sort:
Showing: -Total: -
Page 1 of 1
🔄 Loading latest papers from ArXiv...
Page 1 of 1
Selected Papers in LLM4Rec
A curated collection of seminal papers that have shaped the two main tracks of LLM4Rec: Generative Recommendation and Agentic Recommendation.
🧬 Generative Recommendation
Reformulates recommendation as a sequence generation problem — encoding items as (semantic) IDs or tokens
and training generative/transformer models to directly decode the next item(s) a user will interact with,
replacing the traditional embed-and-retrieve pipeline.
P5 is among the first works to unify diverse recommendation tasks (rating prediction, sequential recommendation, explanation generation, review summarization, direct recommendation) into a single text-to-text generative framework built on a pretrained language model, establishing the "Recommendation as Language Processing" paradigm.
@inproceedings{geng2022recommendation,
title = {Recommendation as Language Processing (RLP): A Unified
Pretrain, Personalized Prompt \& Predict Paradigm (P5)},
author = {Geng, Shijie and Liu, Shuchang and Fu, Zuohui and Ge, Yingqiang and
Zhang, Yongfeng},
booktitle = {Proceedings of the 16th ACM Conference on Recommender Systems},
pages = {299--315},
year = {2022}
}
Generative Rec·Semantic ID & Generative Retrieval·✓ NeurIPS 2023
Shashank Rajput, Nikhil Mehta, Anima Singh, Raghunandan Hulikal Keshavan, Trung Vu, Lukasz Heldt, Lichan Hong, Yi Tay, Vinh Q. Tran, Jonah Samost, Maciej Kula, Ed H. Chi, Maheswaran Sathiamoorthy
TIGER proposes generative retrieval for recommendation: items are mapped to semantically meaningful Semantic IDs via RQ-VAE, and a Transformer sequence-to-sequence model autoregressively decodes the Semantic ID of the next item, replacing nearest-neighbor search over a unified embedding space.
@inproceedings{rajput2023recommender,
title = {Recommender Systems with Generative Retrieval},
author = {Rajput, Shashank and Mehta, Nikhil and Singh, Anima and
Keshavan, Raghunandan Hulikal and Vu, Trung and Heldt, Lukasz and
Hong, Lichan and Tay, Yi and Tran, Vinh Q and Samost, Jonah and
Kula, Maciej and Chi, Ed H and Sathiamoorthy, Maheswaran},
booktitle = {Advances in Neural Information Processing Systems (NeurIPS)},
volume = {36},
year = {2023}
}
Jiaqi Zhai, Lucy Liao, Xing Liu, Yueming Wang, Rui Li, Xuan Cao, Leon Gao, Zhaojie Gong, Fangda Gu, Michael He, Yinghai Lu, Yu Shi
This Meta paper reformulates industrial-scale recommendation as a sequential transduction task and proposes HSTU, a new generative-recommender architecture that scales to trillions of parameters, outperforms FlashAttention2-based Transformers, and empirically follows a power-law scaling trend with training compute.
@inproceedings{zhai2024actions,
title = {Actions Speak Louder than Words: Trillion-Parameter Sequential
Transducers for Generative Recommendations},
author = {Zhai, Jiaqi and Liao, Lucy and Liu, Xing and Wang, Yueming and
Li, Rui and Cao, Xuan and Gao, Leon and Gong, Zhaojie and Gu, Fangda
and He, Michael and Lu, Yinghai and Shi, Yu},
booktitle = {Proceedings of the 41st International Conference on Machine Learning},
year = {2024}
}
Generative Rec·Semantic ID & Generative Retrieval·✓ Industrial Deployment (Kuaishou)
OneRec replaces Kuaishou's cascaded retrieve-and-rank pipeline with a single end-to-end generative model: an encoder-decoder with sparse Mixture-of-Experts that performs session-wise generation, further improved with iterative preference alignment, and is deployed in production on a short-video platform with hundreds of millions of daily active users.
@article{deng2025onerec,
title = {OneRec: Unifying Retrieve and Rank with Generative Recommender
and Iterative Preference Alignment},
author = {Deng, Jiaxin and Wang, Shiyao and Cai, Kuo and Ren, Lejian and
Hu, Qigen and Ding, Weifeng and Luo, Qiang and Zhou, Guorui},
journal = {arXiv preprint arXiv:2502.18965},
year = {2025}
}
🤖 Agentic Recommendation
Treats recommendation as a reasoning and decision-making process carried out by LLM-powered agents —
equipped with memory, tools, planning, and multi-agent collaboration — rather than a single-shot scoring/ranking function.
Agentic Rec·Recommender-oriented·✓ ACL 2026·🌟 Our Work
Weixin Chen, Yuhan Zhao, Jingyuan Huang, Zihe Ye, Clark Mingxuan Ju, Tong Zhao, Neil Shah, Li Chen, Yongfeng Zhang
Agentic MemoryLLMCollaborative Filtering
The evolution of recommender systems has shifted preference storage from rating matrices and dense embeddings to semantic memory in the agentic era...Read more
@inproceedings{chen2026memrec,
title = {MemRec: Collaborative Memory-Augmented Agentic Recommender System},
author = {Chen, Weixin and Zhao, Yuhan and Huang, Jingyuan and Ye, Zihe and
Ju, Clark Mingxuan and Zhao, Tong and Shah, Neil and Chen, Li and
Zhang, Yongfeng},
year = 2026,
booktitle = {Proceedings of the Annual Meeting of the Association for Computational Linguistics (ACL 2026)},
}
Wujiang Xu, Yunxiao Shi, Zujie Liang, Xuying Ning, Kai Mei, Kun Wang, Xi Zhu, Min Xu, Yongfeng Zhang
This work proposes iAgent, an LLM-based agent that acts as a protective shield between users and recommender systems, enabling more transparent and controllable recommendation processes.
@inproceedings{xu2025iagent,
title = {i{A}gent: {LLM} Agent as a Shield between User and Recommender Systems},
author = {Xu, Wujiang and Shi, Yunxiao and Liang, Zujie and Ning, Xuying and
Mei, Kai and Wang, Kun and Zhu, Xi and Xu, Min and Zhang, Yongfeng},
booktitle = {Findings of the Association for Computational Linguistics, {ACL} 2025},
pages = {18056--18084},
year = {2025}
}
Zhefan Wang, Yuanqing Yu, Wendi Zheng, Weizhi Ma, Min Zhang
MACRec introduces a multi-agent collaboration framework where multiple LLM agents work together to improve recommendation quality through collaborative reasoning and information sharing.
@inproceedings{wang2024macrec,
title = {Macrec: A Multi-agent Collaboration Framework for Recommendation},
author = {Wang, Zhefan and Yu, Yuanqing and Zheng, Wendi and Ma, Weizhi and
Zhang, Min},
booktitle = {Proceedings of the 47th International ACM SIGIR Conference on Research
and Development in Information Retrieval},
pages = {2760--2764},
year = {2024}
}
An Zhang, Yuxin Chen, Leheng Sheng, Xiang Wang, Tat-Seng Chua
This paper explores the potential of generative agents powered by large language models in recommendation scenarios, discussing their capabilities, challenges, and future directions.
@inproceedings{zhang2024generative,
title = {On Generative Agents in Recommendation},
author = {Zhang, An and Chen, Yuxin and Sheng, Leheng and Wang, Xiang and
Chua, Tat-Seng},
booktitle = {Proceedings of the 47th International ACM SIGIR Conference on Research
and Development in Information Retrieval},
pages = {1807--1817},
year = {2024}
}
This work presents a framework for simulating realistic user behaviors using LLM-based agents, enabling better testing and evaluation of recommender systems without real user data.
@article{wang2025user,
title = {User Behavior Simulation with Large Language Model-based Agents},
author = {Wang, Lei and Zhang, Jingsen and Yang, Hao and Chen, Zhi-Yuan and
Tang, Jiakai and Zhang, Zeyu and Chen, Xu and Lin, Yankai and Sun, Hao
and Song, Ruihua and others},
journal = {ACM Transactions on Information Systems},
volume = {43},
number = {2},
pages = {1--37},
year = {2025}
}
Yubo Shu, Hansu Gu, Peng Zhang, Haonan Zhang, Tun Lu, Dongsheng Li, Ning Gu
RAH proposes a human-centric framework that leverages LLMs as assistants to facilitate better interaction between users and recommender systems, emphasizing user control and interpretability.
@article{shu2024rah,
title = {{RAH}! {R}ecSys-Assistant-Human: A Human-Centered Recommendation
Framework with {LLM} Agents},
author = {Shu, Yubo and Gu, Hansu and Zhang, Peng and Zhang, Haonan and Lu, Tun
and Li, Dongsheng and Gu, Ning},
journal = {IEEE Transactions on Computational Social Systems},
volume = {11},
number = {5},
pages = {6759--6770},
year = {2024}
}
Yancheng Wang, Ziyan Jiang, Zheng Chen, Fan Yang, Yingxue Zhou, Eunah Cho, Xing Fan, Yanbin Lu, Xiaojiang Huang, Yingzhen Yang
RecMind introduces a comprehensive agent framework powered by large language models that can understand user intent, reason about preferences, and generate personalized recommendations through multi-step reasoning.
@inproceedings{wang2024recmind,
title = {Recmind: Large Language Model Powered Agent for Recommendation},
author = {Wang, Yancheng and Jiang, Ziyan and Chen, Zheng and Yang, Fan and
Zhou, Yingxue and Cho, Eunah and Fan, Xing and Lu, Yanbin and
Huang, Xiaojiang and Yang, Yingzhen},
booktitle = {Findings of the Association for Computational Linguistics: NAACL 2024},
pages = {4351--4364},
year = {2024}
}
InteRecAgent presents an AI agent that integrates large language models to enable interactive and conversational recommendations, supporting natural language queries and explanations for better user engagement.
@article{huang2025recommender,
title = {Recommender Ai Agent: Integrating Large Language Models for Interactive
Recommendations},
author = {Huang, Xu and Lian, Jianxun and Lei, Yuxuan and Yao, Jing and
Lian, Defu and Xie, Xing},
journal = {ACM Transactions on Information Systems},
volume = {43},
number = {4},
pages = {1--33},
year = {2025}
}
Datasets & Resources
Commonly used benchmark datasets and toolkits for LLM4Rec research.
📊 Benchmark Datasets
Standard datasets widely used for evaluating both generative and agentic recommendation methods, including one we've processed and released ourselves.