Elamur Research⌕
☰
home / papers / 2026 / From Personal to Collective: On the Role of Local and Global Knowledge in LLM Personalization

From Personal to Collective: On the Role of Local and Global Knowledge in LLM Personalization

Paper · 2026 · Zehong Wang, Junlin Wu, Zhaoxuan Tan et al.

Summary

Most LLM personalization methods treat every user separately and adapt only to that person's interaction history, using retrieval or per-user adapters such as OPPU. The authors argue that this ignores patterns shared across the user population. The result is weak performance for users with little history (cold start) and overfitting for users whose histories are large but lopsided. They propose LoGo, which combines a local component (a user's own prior interactions) with a global component learned from many users. The global part is updated across successive time phases to follow shifting population preferences. It also groups users into communities with similar interests. A mediator module decides how much weight each source receives, so collective trends inform predictions without overriding individual taste. LoGo works in both parametric (white-box, adapter-based) and non-parametric (black-box, RAG-style) settings. On five LaMP personalization benchmarks it beats existing methods. Accuracy improves for sparse-history users, and output diversity rises for highly active users. These results suggest that collective knowledge should be a core ingredient of LLM personalization.

Key points

  • The evaluation covers five LaMP tasks: LaMP-2N, LaMP-2M, LaMP-3, LaMP-4 and LaMP-5. Code is released at github.com/Zehong-Wang/LoGo.
  • The global knowledge is trained on interaction histories from only 100 users, a limit set by the dataset. The authors also note that experiments use a single model and task setup.
  • Temporal granularity matters: with the default T=5 time splits, performance improves as T grows from 1 to 5 to 10, and T=1 causes a substantial drop. Retrieving more local history items (k from 1 to 2 to 4) also helps steadily.
  • Global parameters stay stable across phases for LaMP-2N, LaMP-2M and LaMP-4, with cosine similarity above 0.85. LaMP-3 and LaMP-5 drift more in later phases.
  • Under temporal shift, a model using only global knowledge degrades in early phases, while the combined local and global model stays robust. The paper flags central aggregation of user histories as a privacy risk and suggests federated learning or differential privacy as mitigations.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

 Institutions

🎓University of Illinois Urbana-ChampaignUrbana · United States209 works · 71,394 citations🎓University of Notre DameNotre Dame · United States59 works · 37,127 citations🎓Purdue University West LafayetteWest Lafayette · United States26 works · 2,076 citations🎓Washington University in St. LouisSt Louis · United States15 works · 1,129 citations🎓Nankai UniversityTianjin · China36 works · 762 citations

 Fields, topics and concepts

📚 Shelves 5The field of science this work belongs to (OpenAlex fields).Computer Science · 18,614 worksArtificial Intelligence · 17,965 worksArts and Humanities · 17,450 worksLanguage and Linguistics · 17,296 worksInformation Systems · 1,730 works

🧭 Topics 1Research subjects and languages the work is about.Language models & pretraining · 3,894 works

Cite

Cite as Zehong Wang, Junlin Wu, Zhaoxuan Tan, Bolian Li, Xianrui Zhong, Zheli Liu and Qingkai Zeng. 2026. From Personal to Collective: On the Role of Local and Global Knowledge in LLM Personalization. Findings of the Association for Computational Linguistics: ACL 2026.

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/2026.findings-acl.1478

@inproceedings{wang2026personal,
  title = {{From Personal to Collective: On the Role of Local and Global Knowledge in LLM Personalization}},
  author = {Zehong Wang, Junlin Wu, Zhaoxuan Tan, Bolian Li, Xianrui Zhong, Zheli Liu and Qingkai Zeng},
  booktitle = {Findings of the Association for Computational Linguistics: ACL 2026},
  year = {2026},
  doi = {10.18653/v1/2026.findings-acl.1478},
  url = {https://doi.org/10.18653/v1/2026.findings-acl.1478}
}

References

49 references, 12 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Aishwarya Gowda AG, Hui-Kai Su, and Wen-Kai Kuo. 2024. Personalized e-commerce: Enhancing customer experience through machine learning-driven personalization. In 2024 IEEE International Conference on Information Technology, Electronics and Intelligent Communication Systems (ICITEICS), pages 1–5. IEE… find
  2. Kemal Bicakci and Nazife Baykal. 2004. Server assisted signatures revisited. In Cryptographers’ Track at the RSA Conference, pages 143–156. Springer. find
  3. Jiangjie Chen, Xintao Wang, Rui Xu, Siyu Yuan, Yikai Zhang, Wei Shi, Jian Xie, Shuang Li, Ruihan Yang, Tinghui Zhu, and 1 others. 2024a. From persona to personalization: A survey on role-playing language agents. arXiv preprint arXiv:2404.18231.
  4. Jin Chen, Zheng Liu, Xu Huang, Chenwang Wu, Qi Liu, Gangwei Jiang, Yuanhao Pu, Yuxuan Lei, Xiaolong Chen, Xingmei Wang, and 1 others. 2024b. When large language models meet personalization: Perspectives of challenges and opportunities. World Wide Web, 27(4):42. find
  5. Pengyu Cheng, Jiawen Xie, Ke Bai, Yong Dai, and Nan Du. 2023. Everyone deserves a reward: Learning customized human preferences. Preprint, arXiv:2309.03126. — [arXiv:2309.03126](https://arxiv.org/abs/2309.03126) find
  6. Yashar Deldjoo, Zhankui He, Julian McAuley, Anton Korikov, Scott Sanner, Arnau Ramisa, René Vidal, Maheswaran Sathiamoorthy, Atoosa Kasirzadeh, and Silvia Milano. 2024. A review of modern recommender systems using generative models (gen-recsys). In Proceedings of the 30th ACM SIGKDD conference on Kn… find
  7. Abhimanyu Dubey, Abhinav Jauhri, Abhinav Pandey, Abhishek Kadian, Ahmad Al-Dahle, Aiesha Letman, Akhil Mathur, Alan Schelten, Amy Yang, Angela Fan, and 1 others. 2024. The llama 3 herd of models. arXiv e-prints, pages arXiv–2407.
  8. Cynthia Dwork. 2006. Differential privacy. In International colloquium on automata, languages, and programming, pages 1–12. Springer. find
  9. Aman Gupta, Anup Shirgaonkar, Angels de Luis Balaguer, Bruno Silva, Daniel Holstein, Dawei Li, Jennifer Marsman, Leonardo O Nunes, Mahsa Rouzbah-man, Morris Sharp, and 1 others. 2024. Rag vs fine-tuning: Pipelines, tradeoffs, and a case study on agriculture. arXiv preprint arXiv:2401.08406. — [arXiv:2401.08406](https://arxiv.org/abs/2401.08406) find
  10. Lin He, Fuchang Li, Haikun Xu, Wenbo Xia, Xuefei Zhang, and Xiaofeng Tao. 2023. Blockchain-based vehicular edge computing networks: the communication perspective. Science China Information Sciences, 66(7):172301. find
  11. Xiangnan He, Lizi Liao, Hanwang Zhang, Liqiang Nie, Xia Hu, and Tat-Seng Chua. 2017. Neural collaborative filtering. In Proceedings of the 26th international conference on world wide web, pages 173–182. find
  12. Michael Hechter and Karl-Dieter Opp. 2001. Social norms. find
  13. Wang-Cheng Kang, Jianmo Ni, Nikhil Mehta, Maheswaran Sathiamoorthy, Lichan Hong, Ed Chi, and Derek Zhiyuan Cheng. 2023. Do llms understand user preferences? evaluating llms on user rating prediction. Preprint, arXiv:2305.06474. — [arXiv:2305.06474](https://arxiv.org/abs/2305.06474) find
  14. Jaehyung Kim and Yiming Yang. 2025. Few-shot personalization of llms with mis-aligned responses. In Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers), pages 11943–11974. find
  15. Young-Ho Kim, Eun Kyoung Choe, Bongshin Lee, and Jinwook Seo. 2019. Understanding personal productivity: How knowledge workers define, evaluate, and reflect on their productivity. In Proceedings of the 2019 CHI Conference on Human Factors in Computing Systems, pages 1–12. find
  16. Cheng Li, Mingyang Zhang, Qiaozhu Mei, Yaqing Wang, Spurthi Amba Hombaiah, Yi Liang, and Michael Bendersky. 2023. Teach llms to personalize– an approach inspired by writing education. arXiv preprint arXiv:2308.07968. — [arXiv:2308.07968](https://arxiv.org/abs/2308.07968) find
  17. Hongwei Li, Dongxiao Liu, Yuanshun Dai, Tom H Luan, and Shui Yu. 2015. Personalized search over encrypted data with efficient and secure updates in mobile clouds. IEEE Transactions on Emerging Topics in Computing, 6(1):97–109. find
  18. Li Li, Yuxi Fan, Mike Tse, and Kuo-Yi Lin. 2020. A review of applications in federated learning. Computers & Industrial Engineering, 149:106854. find
  19. Xiang Lisa Li and Percy Liang. 2021. Prefix-tuning: Optimizing continuous prompts for generation. In Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers), pages 4582–… find
  20. Xinyu Li, Ruiyang Zhou, Zachary C Lipton, and Liu Leqi. 2024. Personalized language modeling from personalized human feedback. arXiv preprint arXiv:2402.05133. — [arXiv:2402.05133](https://arxiv.org/abs/2402.05133) find
  21. Yehuda Lindell. 2020. Secure multiparty computation. Communications of the ACM, 64(1):86–96. find
  22. Edward J Hu, Phillip Wallis, Zeyuan Allen-Zhu, Yuanzhi Li, Shean Wang, Lu Wang, Weizhu Chen, and 1 others. 2021. Lora: Low-rank adaptation of large language models. In International Conference on Learning Representations.
  23. Jiongnan Liu, Yutao Zhu, Shuting Wang, Xiaochi Wei, Erxue Min, Yu Lu, Shuaiqiang Wang, Dawei Yin, and Zhicheng Dou. 2025. LLMs + persona-plug = personalized LLMs. In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (ACL).
  24. Joel Jang, Seungone Kim, Bill Yuchen Lin, Yizhong Wang, Jack Hessel, Luke Zettlemoyer, Hannaneh Hajishirzi, Yejin Choi, and Prithviraj Ammanabrolu. 2023. Personalized soups: Personalized large language model alignment via post-hoc parameter merging. arXiv preprint arXiv:2310.11564.
  25. Sheshera Mysore, Zhuoran Lu, Mengting Wan, Longqi Yang, Steve Menezes, Tina Baghaee, Emmanuel Barajas Gonzalez, Jennifer Neville, and Tara Safavi. 2023. Pearl: Personalizing large language model writing assistants with generation-calibrated retrievers. arXiv preprint arXiv:2311.09180. find
  26. Charles Packer, Vivian Fang, Shishir_G Patil, Kevin Lin, Sarah Wooders, and Joseph_E Gonzalez. 2023. Memgpt: Towards llms as operating systems. find
  27. Chanwoo Park, Mingyang Liu, Dingwen Kong, Kaiqing Zhang, and Asuman E Ozdaglar. 2024. Rlhf from heterogeneous feedback via personalization and preference aggregation. In ICML 2024 Workshop: Aligning Reinforcement Learning Experimentalists and Theorists. find
  28. Mustafa Ayobami Raji, Hameedat Bukola Olodo, Timothy Tolulope Oke, Wilhelmina Afua Addy, Onyeka Chrisanctus Ofodile, and Adedoyin Tolulope Oyewole. 2024. E-commerce and consumer behavior: A review of ai-powered personalization and market trends. GSC advanced research and reviews, 18(3):066–077. find
  29. Chris Richardson, Yao Zhang, Kellen Gillespie, Sudipta Kar, Arshdeep Singh, Zeynab Raeesy, Omar Zia Khan, and Abhinav Sethy. 2023. Integrating summarization and retrieval for enhanced personalization via large language models. arXiv preprint arXiv:2310.20081. language models through collaborative ef… — [arXiv:2310.20081](https://arxiv.org/abs/2310.20081) find
  30. Zhaoxuan Tan, Qingkai Zeng, Yijun Tian, Zheyuan Liu, Bing Yin, and Meng Jiang. 2024b. Democratizing large language models via personalized parameter-efficient fine-tuning. In EMNLP.
  31. Zhaoxuan Tan, Zinan Zeng, Qingkai Zeng, Zhenyu Wu, Zheyuan Liu, Fengran Mo, and Meng Jiang. 2025a. Can large language models understand preferences in personalized recommendation? arXiv preprint arXiv:2501.13391. — [arXiv:2501.13391](https://arxiv.org/abs/2501.13391) find
  32. Zhaoxuan Tan, Zixuan Zhang, Haoyang Wen, Zheng Li, Rongzhi Zhang, Pei Chen, Fengran Mo, Zheyuan Liu, Qingkai Zeng, Qingyu Yin, and 1 others. 2025b. Instant personalized large language model adaptation via hypernetwork. arXiv preprint arXiv:2510.16282.
  33. Leonard Tetzlaff, Florian Schmiedek, and Garvin Brod. 2021. Developing personalized education: A dynamic framework. Educational Psychology Review, 33(3):863–882. find
  34. Stephen Robertson, Hugo Zaragoza, and 1 others. 2009. The probabilistic relevance framework: Bm25 and beyond. Foundations and Trends® in Information Retrieval, 3(4):333–389. find
  35. Valerie Trifts and Hamed Aghakhani. 2019. Enhancing digital entertainment through personalization: The evolving role of product placements. Journal of Marketing Communications, 25(6):607–625. find
  36. Alireza Salemi, Surya Kallumadi, and Hamed Zamani. 2024. Optimization methods for personalizing large language models through retrieval augmentation. In Proceedings of the 47th International ACM SIGIR Conference on Research and Development in Information Retrieval, pages 752–762. find
  37. Bing Wang, Xinnian Liang, Jian Yang, Hui Huang, Shuangzhi Wu, Peihao Wu, Lu Lu, Zejun Ma, and Zhoujun Li. 2023a. Enhancing large language model with self-controlled memory framework. arXiv preprint arXiv:2304.13343. — [arXiv:2304.13343](https://arxiv.org/abs/2304.13343) find
  38. Alireza Salemi, Sheshera Mysore, Michael Bendersky, and Hamed Zamani. 2023. Lamp: When large language models meet personalization. arXiv preprint arXiv:2304.11406.
  39. Danqing Wang, Kevin Yang, Hanlin Zhu, Xiaomeng Yang, Andrew Cohen, Lei Li, and Yuandong Tian. 2023b. Learning personalized story evaluation. arXiv preprint arXiv:2310.03304. — [arXiv:2310.03304](https://arxiv.org/abs/2310.03304) find
  40. J Ben Schafer, Dan Frankowski, Jon Herlocker, and Shilad Sen. 2007. Collaborative filtering recommender systems. In The adaptive web: methods and strategies of web personalization, pages 291–324. Springer. find
  41. Tianxin Wei, Noveen Sachdeva, Benjamin Coleman, Zhankui He, Yuanchen Bei, Xuying Ning, Mengting Ai, Yunzhe Li, Jingrui He, Ed H Chi, and 1 others. 2025. Evo-memory: Benchmarking llm agent test-time learning with self-evolving memory. arXiv preprint arXiv:2511.20857. — [arXiv:2511.20857](https://arxiv.org/abs/2511.20857) find
  42. Chenkai Sun, Ke Yang, Revanth Gangi Reddy, Yi Fung, Hou Pong Chan, Kevin Small, ChengXiang Zhai, and Heng Ji. 2025. Persona-db: Efficient large language model personalization for response prediction with collaborative data refinement. In Proceedings of the 31st International Conference on Computatio…
  43. Le Wu, Yong Ge, Qi Liu, Enhong Chen, Richang Hong, Junping Du, and Meng Wang. 2017. Modeling the evolution of users’ preferences and social links in social networking services. IEEE Transactions on Knowledge and Data Engineering, 29(6):1240–1253. find
  44. Chenkai Sun, Ke Yang, Revanth Gangi Reddy, Yi R Fung, Hou Pong Chan, ChengXiang Zhai, and Heng Ji. 2024. Persona-db: Efficient large language model personalization for response prediction with collaborative data refinement. arXiv preprint arXiv:2402.11060.
  45. Zhaoxuan Tan, Zheyuan Liu, and Meng Jiang. 2024a. Personalized pieces: Efficient personalized large Wujiang Xu, Zujie Liang, Kai Mei, Hang Gao, Juntao Tan, and Yongfeng Zhang. 2025. A-mem: Agentic memory for llm agents. arXiv preprint arXiv:2502.12110.
  46. Wayne Xin Zhao, Kun Zhou, Junyi Li, Tianyi Tang, Xiaolei Wang, Yupeng Hou, Yingqian Min, Beichen Zhang, Junjie Zhang, Zican Dong, and 1 others. 2023. A survey of large language models. arXiv preprint arXiv:2303.18223, 1(2).
  47. Yu Zheng, Chen Gao, Xiang Li, Xiangnan He, Yong Li, and Depeng Jin. 2021. Disentangling user interest and conformity for recommendation with causal embedding. In Proceedings of the web conference 2021, pages 2980–2991. find
  48. Aakas Zhiyuli, Yanfang Chen, Xuan Zhang, and Xun Liang. 2023. Bookgpt: A general framework for book recommendation empowered by large language model. arXiv preprint arXiv:2305.15673. — [arXiv:2305.15673](https://arxiv.org/abs/2305.15673) find
  49. Yuchen Zhuang, Haotian Sun, Yue Yu, Rushi Qiang, Qifan Wang, Chao Zhang, and Bo Dai. 2024. Hydra: Model factorization framework for black-box llm personalization. Advances in Neural Information Processing Systems, 37:100783–100815. find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/wang-2026-personal-collective-role-local-global/