---
title: "Personalized Text Generation with Contrastive Activation Steering"
description: "The authors argue that the two dominant approaches have problems. 44 references, cited by 9 papers here."
authors: ["Jinghao Zhang", "Yuting Liu", "Wenjie Wang", "Qiang Liu", "Shu Wu", "Liang Wang", "Tat-Seng Chua"]
year: 2025
venue: "Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)"
source: "https://aclanthology.org/2025.acl-long.353/"
doi: "10.18653/v1/2025.acl-long.353"
licence: "CC BY 4.0 (ACL Anthology)"
licence_url: "https://creativecommons.org/licenses/by/4.0/"
summary_licence: "Elamur Cosmic License"
canonical: "https://research.elamur.ai/free/en/papers/zhang-2025-personalized-text-generation-contrastive-activation/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/zhang-2025-personalized-text-generation-contrastive-activation/)

[home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2025](https://md.research.elamur.ai/free/en/papers/2025/page/7/) / Personalized Text Generation with Contrastive Activation Steering

# Personalized Text Generation with Contrastive Activation Steering

**Authors:** [Jinghao Zhang](https://md.research.elamur.ai/free/en/authors/jinghao-zhang/), Yuting Liu, [Wenjie Wang](https://md.research.elamur.ai/free/en/authors/wenjie-wang/), [Qiang Liu](https://md.research.elamur.ai/free/en/authors/qiang-liu/), [Shu Wu](https://md.research.elamur.ai/free/en/authors/shu-wu/), [Liang Wang](https://md.research.elamur.ai/free/en/authors/liang-wang/) and [Tat-Seng Chua](https://md.research.elamur.ai/free/en/authors/tat-seng-chua/) · **Year:** 2025 · **Venue:** [Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)](https://md.research.elamur.ai/free/en/venues/acl/) ([ACL 2025](https://md.research.elamur.ai/free/en/venues/acl/2025/)) · **Publisher:** [Association for Computational Linguistics](https://md.research.elamur.ai/free/en/publishers/association-for-computational-linguistics/) · **Licence:** [CC BY 4.0 (ACL Anthology)](https://creativecommons.org/licenses/by/4.0/)

**Links:** [Publisher's page](https://aclanthology.org/2025.acl-long.353/) · [PDF](https://aclanthology.org/2025.acl-long.353.pdf) · [DOI 10.18653/v1/2025.acl-long.353](https://doi.org/10.18653/v1/2025.acl-long.353) · [OpenAlex](https://openalex.org/W4412945555)

**Citations:** 4 citations (OpenAlex, 2026-10-14) · [9](https://md.research.elamur.ai/free/en/papers/zhang-2025-personalized-text-generation-contrastive-activation/cited-by/) in our library, of the 149,439 papers whose bibliographies we read

## Summary
The paper addresses personalized text generation: producing outputs, such as emails or headlines, that match an individual user's writing style as shown in their past texts. The authors argue that the two dominant approaches have problems. RAG and per-user PEFT both treat past texts as wholes, so topic and style stay mixed. Both also scale poorly: RAG adds retrieval latency, and PEFT needs a stored adapter for every user. Their framework, StyleVector, requires no training. A general LLM first writes neutral answers to the user's past prompts. The method then compares the model's hidden activations for the user's real answers against those for the neutral ones. The resulting direction serves as a single vector that captures the user's style. At inference, this vector is added to the activations at a chosen layer to push generation toward that style. On the LaMP (short-form) and LongLaMP (long-form) benchmarks, StyleVector improves personalization by about 8% relative and needs roughly 1700× less storage than PEFT. This makes per-user personalization cheaper to deploy at scale.

## Key points
- The authors test three ways to derive the style vector from last-token hidden states at a layer ℓ: mean difference of the paired activations, a normalized logistic regression direction, and the first principal component of the activation differences.
- Building a user's vector costs only 2|P_u| forward passes and no back-propagation. Only one vector of size O(D) is stored per user, compared with O(rDL) for LoRA-style adapters.
- The model that writes the neutral responses can be any open or closed LLM. The models used include LLaMA-2, Contriver for retrieval baselines, and the GPT-3.5-turbo and DeepSeek-Chat APIs.
- The headline results are about 8% relative gain in personalization quality and about 1700× lower storage than PEFT, along with lower latency than RAG.
- Stated limitations:
  - Style and content may be only imperfectly disentangled.
  - A single vector may merge several distinct style dimensions.
  - LaMP and LongLaMP assume each user writes in one domain, so cross-domain style consistency remains untested.

*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

- [National University of Singapore](https://md.research.elamur.ai/free/en/universities/national-university-of-singapore/) — University, Singapore, Singapore · 180 works · 34,304 citations
- [University of Science and Technology of China](https://md.research.elamur.ai/free/en/universities/university-of-science-and-technology-of-china/) — University, Hefei, China · 108 works · 112,698 citations
- [Chinese Academy of Sciences](https://md.research.elamur.ai/free/en/government/chinese-academy-of-sciences/) — Government body, Beijing, China · 194 works · 23,906 citations
- [Institute of Automation](https://md.research.elamur.ai/free/en/labs/institute-of-automation/) — Research institute, Beijing, China · 70 works · 14,251 citations
- [University of Chinese Academy of Sciences](https://md.research.elamur.ai/free/en/universities/university-of-chinese-academy-of-sciences/) — University, Beijing, China · 147 works · 8,480 citations
- [Northeastern University](https://md.research.elamur.ai/free/en/universities/northeastern-university-cn/) — University, Shenyang, China · 42 works · 1,940 citations

📚 Shelves (5): [🖥️ Computer Science](https://md.research.elamur.ai/free/en/shelves/computer-science/) · 16,667 works · [🤖 Artificial Intelligence](https://md.research.elamur.ai/free/en/shelves/computer-science/artificial-intelligence/) · 16,194 works · [🎭 Arts and Humanities](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/) · 16,039 works · [🗣️ Language and Linguistics](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/language-and-linguistics/) · 15,891 works · [👁️ Computer Vision and Pattern Recognition](https://md.research.elamur.ai/free/en/shelves/computer-science/computer-vision-and-pattern-recognition/) · 2,869 works

🧭 Topics (1): [✍️ Text generation](https://md.research.elamur.ai/free/en/topics/text-generation/) · 202 works

## Cite

Cite as: Jinghao Zhang, Yuting Liu, Wenjie Wang, Qiang Liu, Shu Wu, Liang Wang and Tat-Seng Chua. 2025. Personalized Text Generation with Contrastive Activation Steering. Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers).

The paper itself, not this page. BibTeX from the library's record of the work.

```bibtex
@inproceedings{zhang2025personalized,
  title = {{Personalized Text Generation with Contrastive Activation Steering}},
  author = {Jinghao Zhang, Yuting Liu, Wenjie Wang, Qiang Liu, Shu Wu, Liang Wang and Tat-Seng Chua},
  booktitle = {Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)},
  year = {2025},
  doi = {10.18653/v1/2025.acl-long.353},
  url = {https://doi.org/10.18653/v1/2025.acl-long.353}
}
```

- [← previous Pangu Embedded: An Efficient Dual-system LLM Reasoner with Metacognition](https://md.research.elamur.ai/free/en/papers/chen-2025-pangu-embedded-efficient-dual/)
- [next → POSITION BIAS MITIGATES POSITION BIAS: Mitigate Position Bias Through Inter-Position Knowledge Distillation](https://md.research.elamur.ai/free/en/papers/wang-2025-position-bias-mitigates-position-bias/)
- [↑ back to papers of 2025](https://md.research.elamur.ai/free/en/papers/2025/page/7/)

## References

44 references, 14 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Bashar Alhafni, Vivek Kulkarni, Dhruv Kumar, and Vipul Raheja. 2024. Personalized text generation with fine-grained linguistic control. ArXiv. [find](https://scholar.google.com/scholar?q=Bashar+Alhafni%2C+Vivek+Kulkarni%2C+Dhruv+Kumar%2C+and+Vipul+Raheja%2E+2024%2E+Personalized+text+generation+with+fine%2Dgrained+linguistic+control%2E+ArXiv%2E)
2. Steven Au, Cameron J Dimacali, Ojasmitha Pedirappa-gari, Namyong Park, Franck Dernoncourt, Yu Wang, Nikos Kanakaris, Hanieh Deilamsalehy, Ryan A Rossi, and Nesreen K Ahmed. 2025. Personalized graph-based retrieval for large language models. arXiv preprint arXiv:2501.02157. [find](https://scholar.google.com/scholar?q=Steven+Au%2C+Cameron+J+Dimacali%2C+Ojasmitha+Pedirappa%2Dgari%2C+Namyong+Park%2C+Franck+Dernoncourt%2C+Yu+Wang%2C+Nikos+Kanakaris%2C+Hanieh+Deilamsalehy%2C+Ryan+A+Rossi%2C+and+Nesreen+K+Ahmed%2E+2025%2E+Personalized+graph%2Dba)
3. Hongru Cai, Yongqi Li, Wenjie Wang, Fengbin Zhu, Xiaoyu Shen, Wenjie Li, and Tat-Seng Chua. 2024. Large language models empowered personalized web agents. arXiv preprint arXiv:2410.17236. — [arXiv:2410.17236](https://arxiv.org/abs/2410.17236) [find](https://scholar.google.com/scholar?q=Hongru+Cai%2C+Yongqi+Li%2C+Wenjie+Wang%2C+Fengbin+Zhu%2C+Xiaoyu+Shen%2C+Wenjie+Li%2C+and+Tat%2DSeng+Chua%2E+2024%2E+Large+language+models+empowered+personalized+web+agents%2E+arXiv+preprint+arXiv%3A2410%2E17236%2E+%E2%80%94+%5BarXiv%3A241)
4. Zhongzhi Chen, Xingwu Sun, Xianfeng Jiao, Fengzong Lian, Zhanhui Kang, Di Wang, and Chengzhong Xu. 2024. Truth forest: Toward multi-scale truthfulness in large language models through intervention without tuning. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 38, pages 2096… [find](https://scholar.google.com/scholar?q=Zhongzhi+Chen%2C+Xingwu+Sun%2C+Xianfeng+Jiao%2C+Fengzong+Lian%2C+Zhanhui+Kang%2C+Di+Wang%2C+and+Chengzhong+Xu%2E+2024%2E+Truth+forest%3A+Toward+multi%2Dscale+truthfulness+in+large+language+models+through+intervention+wit)
5. Ziyang Chen and Stylios Moscholios. 2024. Using prompts to guide large language models in imitating a real person’s language style. ArXiv. [find](https://scholar.google.com/scholar?q=Ziyang+Chen+and+Stylios+Moscholios%2E+2024%2E+Using+prompts+to+guide+large+language+models+in+imitating+a+real+person%E2%80%99s+language+style%2E+ArXiv%2E)
6. Hoagy Cunningham, Aidan Ewart, Logan Riggs, Robert Huben, and Lee Sharkey. 2023. Sparse autoencoders find highly interpretable features in language models. arXiv preprint arXiv:2309.08600. [find](https://scholar.google.com/scholar?q=Hoagy+Cunningham%2C+Aidan+Ewart%2C+Logan+Riggs%2C+Robert+Huben%2C+and+Lee+Sharkey%2E+2023%2E+Sparse+autoencoders+find+highly+interpretable+features+in+language+models%2E+arXiv+preprint+arXiv%3A2309%2E08600%2E)
7. Boyi Deng, Wenjie Wang, Fengbin Zhu, Qifan Wang, and Fuli Feng. 2025. Cram: Credibility-aware attention modification in llms for combating misinformation in rag. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 39, pages 23760– 23768. [find](https://scholar.google.com/scholar?q=Boyi+Deng%2C+Wenjie+Wang%2C+Fengbin+Zhu%2C+Qifan+Wang%2C+and+Fuli+Feng%2E+2025%2E+Cram%3A+Credibility%2Daware+attention+modification+in+llms+for+combating+misinformation+in+rag%2E+In+Proceedings+of+the+AAAI+Conference+)
8. Yucheng Ding, Yangwenjian Tan, Xiangyu Liu, Chaoyue Niu, Fandong Meng, Jie Zhou, Ning Liu, Fan Wu, and Guihai Chen. 2025. Personalized language model learning on text data without user identifiers. ArXiv. [find](https://scholar.google.com/scholar?q=Yucheng+Ding%2C+Yangwenjian+Tan%2C+Xiangyu+Liu%2C+Chaoyue+Niu%2C+Fandong+Meng%2C+Jie+Zhou%2C+Ning+Liu%2C+Fan+Wu%2C+and+Guihai+Chen%2E+2025%2E+Personalized+language+model+learning+on+text+data+without+user+identifiers%2E+Ar)
9. Jillian R. Fisher, Skyler Hallinan, Ximing Lu, Mitchell Gordon, Zaid Harchaoui, and Yejin Choi. 2024. [Styleremix: Interpretable authorship obfuscation via distillation and perturbation of style elements](https://md.research.elamur.ai/free/en/papers/fisher-2024-styleremix-interpretable-authorship-obfuscation-distillation/). ArXiv, abs/2408.15666.
10. Edward J Hu, Yelong Shen, Phillip Wallis, Zeyuan Allen-Zhu, Yuanzhi Li, Shean Wang, Lu Wang, and Weizhu Chen. 2021. [Lora: Low-rank adaptation of large language models](https://md.research.elamur.ai/free/en/papers/hu-2021-lora-low-rank-adaptation-large/). arXiv preprint arXiv:2106.09685.
11. 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. [find](https://scholar.google.com/scholar?q=Joel+Jang%2C+Seungone+Kim%2C+Bill+Yuchen+Lin%2C+Yizhong+Wang%2C+Jack+Hessel%2C+Luke+Zettlemoyer%2C+Hannaneh+Hajishirzi%2C+Yejin+Choi%2C+and+Prithviraj+Ammanabrolu%2E+2023%2E+Personalized+soups%3A+Personalized+large+languag)
12. Ishita Kumar, Snigdha Viswanathan, Sushrita Yerra, Alireza Salemi, Ryan A Rossi, Franck Dernoncourt, Hanieh Deilamsalehy, Xiang Chen, Ruiyi Zhang, Shubham Agarwal, et al. 2024. Longlamp: A benchmark for personalized long-form text generation. arXiv preprint arXiv:2407.11016. — [arXiv:2407.11016](https://arxiv.org/abs/2407.11016) [find](https://scholar.google.com/scholar?q=Ishita+Kumar%2C+Snigdha+Viswanathan%2C+Sushrita+Yerra%2C+Alireza+Salemi%2C+Ryan+A+Rossi%2C+Franck+Dernoncourt%2C+Hanieh+Deilamsalehy%2C+Xiang+Chen%2C+Ruiyi+Zhang%2C+Shubham+Agarwal%2C+et+al%2E+2024%2E+Longlamp%3A+A+benchmark+f)
13. Yibin Lei, Liang Ding, Yu Cao, Changtong Zan, Andrew Yates, and Dacheng Tao. 2023. Unsupervised dense retrieval with relevance-aware contrastive pre-training. In Findings of the Association for Computational Linguistics: ACL 2023, pages 10932–10940. [find](https://scholar.google.com/scholar?q=Yibin+Lei%2C+Liang+Ding%2C+Yu+Cao%2C+Changtong+Zan%2C+Andrew+Yates%2C+and+Dacheng+Tao%2E+2023%2E+Unsupervised+dense+retrieval+with+relevance%2Daware+contrastive+pre%2Dtraining%2E+In+Findings+of+the+Association+for+Comput)
14. Tom Lieberum, Senthooran Rajamanoharan, Arthur Conmy, Lewis Smith, Nicolas Sonnerat, Vikrant Varma, János Kramár, Anca Dragan, Rohin Shah, and Neel Nanda. 2024. Gemma scope: Open sparse autoencoders everywhere all at once on gemma 2. arXiv preprint arXiv:2408.05147. [find](https://scholar.google.com/scholar?q=Tom+Lieberum%2C+Senthooran+Rajamanoharan%2C+Arthur+Conmy%2C+Lewis+Smith%2C+Nicolas+Sonnerat%2C+Vikrant+Varma%2C+J%C3%A1nos+Kram%C3%A1r%2C+Anca+Dragan%2C+Rohin+Shah%2C+and+Neel+Nanda%2E+2024%2E+Gemma+scope%3A+Open+sparse+autoencoders+e)
15. Zihao Lin, Zichao Wang, Yuanting Pan, Varun Manjunatha, Ryan Rossi, Angela Lau, Lifu Huang, and Tong Sun. 2024. Persona-sq: A personalized suggested question generation framework for real-world documents. arXiv preprint arXiv:2412.12445. [find](https://scholar.google.com/scholar?q=Zihao+Lin%2C+Zichao+Wang%2C+Yuanting+Pan%2C+Varun+Manjunatha%2C+Ryan+Rossi%2C+Angela+Lau%2C+Lifu+Huang%2C+and+Tong+Sun%2E+2024%2E+Persona%2Dsq%3A+A+personalized+suggested+question+generation+framework+for+real%2Dworld+docume)
16. Jiongnan Liu, Yutao Zhu, Shuting Wang, Xiaochi Wei, Erxue Min, Yu Lu, Shuaiqiang Wang, Dawei Yin, and Zhicheng Dou. 2024. [Llms + persona-plug = personalized llms](https://md.research.elamur.ai/free/en/papers/liu-2025-llms-persona-plug-personalized-llms/). ArXiv.
17. Sheng Liu, Haotian Ye, Lei Xing, and James Zou. 2023. In-context vectors: Making in context learning more effective and controllable through latent space steering. arXiv preprint arXiv:2311.06668. [find](https://scholar.google.com/scholar?q=Sheng+Liu%2C+Haotian+Ye%2C+Lei+Xing%2C+and+James+Zou%2E+2023%2E+In%2Dcontext+vectors%3A+Making+in+context+learning+more+effective+and+controllable+through+latent+space+steering%2E+arXiv+preprint+arXiv%3A2311%2E06668%2E)
18. Yuting Liu, Jinghao Zhang, Yizhou Dang, Yuliang Liang, Qiang Liu, Guibing Guo, Jianzhe Zhao, and Xingwei Wang. 2025. Cora: Collaborative information perception by large language model’s weights for recommendation. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 39, pages 122… [find](https://scholar.google.com/scholar?q=Yuting+Liu%2C+Jinghao+Zhang%2C+Yizhou+Dang%2C+Yuliang+Liang%2C+Qiang+Liu%2C+Guibing+Guo%2C+Jianzhe+Zhao%2C+and+Xingwei+Wang%2E+2025%2E+Cora%3A+Collaborative+information+perception+by+large+language+model%E2%80%99s+weights+for+re)
19. Zheqi Lv, Tianyu Zhan, Wenjie Wang, Xinyu Lin, Shengyu Zhang, Wenqiao Zhang, Jiwei Li, Kun Kuang, and Fei Wu. 2025. Collaboration of large language models and small recommendation models for device-cloud recommendation. In Proceedings of the 31st ACM SIGKDD Conference on Knowledge Discovery and Data… [find](https://scholar.google.com/scholar?q=Zheqi+Lv%2C+Tianyu+Zhan%2C+Wenjie+Wang%2C+Xinyu+Lin%2C+Shengyu+Zhang%2C+Wenqiao+Zhang%2C+Jiwei+Li%2C+Kun+Kuang%2C+and+Fei+Wu%2E+2025%2E+Collaboration+of+large+language+models+and+small+recommendation+models+for+device%2Dcl)
20. Adam Paszke, Sam Gross, Francisco Massa, Adam Lerer, James Bradbury, Gregory Chanan, Trevor Killeen, Zeming Lin, Natalia Gimelshein, Luca Antiga, et al. 2019. [Pytorch: An imperative style, high-performance deep learning library](https://md.research.elamur.ai/free/en/papers/paszke-2019-pytorch-imperative-style-high-performance/). Advances in neural information processing systems, 32.
21. Rafael Rafailov, Archit Sharma, Eric Mitchell, Christopher D Manning, Stefano Ermon, and Chelsea Finn. 2024. [Direct preference optimization: Your language model is secretly a reward model](https://md.research.elamur.ai/free/en/papers/rafailov-2023-direct-preference-optimization-language-model/). Advances in Neural Information Processing Systems, 36.
22. 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. [find](https://scholar.google.com/scholar?q=Chris+Richardson%2C+Yao+Zhang%2C+Kellen+Gillespie%2C+Sudipta+Kar%2C+Arshdeep+Singh%2C+Zeynab+Raeesy%2C+Omar+Zia+Khan%2C+and+Abhinav+Sethy%2E+2023%2E+Integrating+summarization+and+retrieval+for+enhanced+personalization+)
23. Nina Rimsky, Nick Gabrieli, Julian Schulz, Meg Tong, Evan Hubinger, and Alexander Turner. 2024. Steering llama 2 via contrastive activation addition. In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 15504–15522, Bangkok, Thaila… [find](https://scholar.google.com/scholar?q=Nina+Rimsky%2C+Nick+Gabrieli%2C+Julian+Schulz%2C+Meg+Tong%2C+Evan+Hubinger%2C+and+Alexander+Turner%2E+2024%2E+Steering+llama+2+via+contrastive+activation+addition%2E+In+Proceedings+of+the+62nd+Annual+Meeting+of+the+A)
24. Stephen Robertson, Hugo Zaragoza, et al. 2009. The probabilistic relevance framework: Bm25 and beyond. Foundations and Trends® in Information Retrieval, 3(4):333–389. [find](https://scholar.google.com/scholar?q=Stephen+Robertson%2C+Hugo+Zaragoza%2C+et+al%2E+2009%2E+The+probabilistic+relevance+framework%3A+Bm25+and+beyond%2E+Foundations+and+Trends%C2%AE+in+Information+Retrieval%2C+3%284%29%3A333%E2%80%93389%2E)
25. Alireza Salemi, Surya Kallumadi, and Hamed Zamani. 2024a. Optimization methods for personalizing large language models through retrieval augmentation. In Annual International ACM SIGIR Conference on Research and Development in Information Retrieval. [find](https://scholar.google.com/scholar?q=Alireza+Salemi%2C+Surya+Kallumadi%2C+and+Hamed+Zamani%2E+2024a%2E+Optimization+methods+for+personalizing+large+language+models+through+retrieval+augmentation%2E+In+Annual+International+ACM+SIGIR+Conference+on+R)
26. Alireza Salemi, Sheshera Mysore, Michael Bendersky, and Hamed Zamani. 2024b. [LaMP: When large language models meet personalization](https://md.research.elamur.ai/free/en/papers/salemi-2024-lamp-when-large-language-models/). In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 7370–7392, Bangkok, Thailand. Association for…
27. Alireza Salemi and Hamed Zamani. 2024a. Comparing retrieval-augmentation and parameter-efficient fine-tuning for privacy-preserving personalization of large language models. arXiv preprint arXiv:2409.09510. — [arXiv:2409.09510](https://arxiv.org/abs/2409.09510) [find](https://scholar.google.com/scholar?q=Alireza+Salemi+and+Hamed+Zamani%2E+2024a%2E+Comparing+retrieval%2Daugmentation+and+parameter%2Defficient+fine%2Dtuning+for+privacy%2Dpreserving+personalization+of+large+language+models%2E+arXiv+preprint+arXiv%3A2409%2E)
28. Alireza Salemi and Hamed Zamani. 2024b. Learning to rank for multiple retrieval-augmented models through iterative utility maximization. arXiv preprint arXiv:2410.09942. — [arXiv:2410.09942](https://arxiv.org/abs/2410.09942) [find](https://scholar.google.com/scholar?q=Alireza+Salemi+and+Hamed+Zamani%2E+2024b%2E+Learning+to+rank+for+multiple+retrieval%2Daugmented+models+through+iterative+utility+maximization%2E+arXiv+preprint+arXiv%3A2410%2E09942%2E+%E2%80%94+%5BarXiv%3A2410%2E09942%5D%28https%3A%2F%2Fa)
29. Zhaoxuan Tan, Zheyuan Liu, and Meng Jiang. 2024a. [Personalized pieces: Efficient personalized large language models through collaborative efforts](https://md.research.elamur.ai/free/en/papers/tan-2024-personalized-pieces-efficient-personalized-large/). arXiv preprint arXiv:2406.10471.
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](https://md.research.elamur.ai/free/en/papers/tan-2024-democratizing-large-language-models-personalized/). In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing.
31. Zhaoxuan Tan, Zinan Zeng, Qingkai Zeng, Zhenyu Wu, Zheyuan Liu, Fengran Mo, and Meng Jiang. 2025. Can large language models understand preferences in personalized recommendation? ArXiv. [find](https://scholar.google.com/scholar?q=Zhaoxuan+Tan%2C+Zinan+Zeng%2C+Qingkai+Zeng%2C+Zhenyu+Wu%2C+Zheyuan+Liu%2C+Fengran+Mo%2C+and+Meng+Jiang%2E+2025%2E+Can+large+language+models+understand+preferences+in+personalized+recommendation%3F+ArXiv%2E)
32. Hugo Touvron, Louis Martin, Kevin Stone, Peter Albert, Amjad Almahairi, Yasmine Babaei, Nikolay Bashlykov, Soumya Batra, Prajjwal Bhargava, Shruti Bhosale, et al. 2023. [Llama 2: Open foundation and fine-tuned chat models](https://md.research.elamur.ai/free/en/papers/touvron-2023-llama-2-open-foundation-fine/). arXiv preprint arXiv:2307.09288.
33. Andy Zou, Long Phan, Sarah Chen, James Campbell, Phillip Guo, Richard Ren, Alexander Pan, Xuwang Yin, Mantas Mazeika, Ann-Kathrin Dombrowski, et al. 2023. [Representation engineering: A top-down approach to ai transparency](https://md.research.elamur.ai/free/en/papers/zou-2023-representation-engineering-top-down-approach/). arXiv preprint arXiv:2310.01405.
34. Alexander Matt Turner, Lisa Thiergart, Gavin Leech, David Udell, Juan J Vazquez, Ulisse Mini, and Monte MacDiarmid. 2023. Activation addition: Steering language models without optimization. arXiv eprints, pages arXiv–2308. [find](https://scholar.google.com/scholar?q=Alexander+Matt+Turner%2C+Lisa+Thiergart%2C+Gavin+Leech%2C+David+Udell%2C+Juan+J+Vazquez%2C+Ulisse+Mini%2C+and+Monte+MacDiarmid%2E+2023%2E+Activation+addition%3A+Steering+language+models+without+optimization%2E+arXiv+epri)
35. T Wolf. 2019. [Huggingface’s transformers: State-ofthe-art natural language processing](https://md.research.elamur.ai/free/en/papers/wolf-2020-transformers-state-art-natural-language/). arXiv preprint arXiv:1910.03771.
36. Yiyan Xu, Jinghao Zhang, Alireza Salemi, Xinting Hu, Wenjie Wang, Fuli Feng, Hamed Zamani, Xiangnan He, and Tat-Seng Chua. 2025. [Personalized generation in large model era: A survey](https://md.research.elamur.ai/free/en/papers/xu-2025-personalized-generation-large-model-era/). arXiv preprint arXiv:2503.02614.
37. Jiarui Zhang. 2024. Guided profile generation improves personalization with large language models. In Findings of the Association for Computational Linguistics: EMNLP 2024. [find](https://scholar.google.com/scholar?q=Jiarui+Zhang%2E+2024%2E+Guided+profile+generation+improves+personalization+with+large+language+models%2E+In+Findings+of+the+Association+for+Computational+Linguistics%3A+EMNLP+2024%2E)
38. Jinghao Zhang, Yuting Liu, Qiang Liu, Shu Wu, Guibing Guo, and Liang Wang. 2024a. Stealthy attack on large language model based recommendation. In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 5839–5857. [find](https://scholar.google.com/scholar?q=Jinghao+Zhang%2C+Yuting+Liu%2C+Qiang+Liu%2C+Shu+Wu%2C+Guibing+Guo%2C+and+Liang+Wang%2E+2024a%2E+Stealthy+attack+on+large+language+model+based+recommendation%2E+In+Proceedings+of+the+62nd+Annual+Meeting+of+the+Associa)
39. Kai Zhang, Fubang Zhao, Yangyang Kang, and Xiaozhong Liu. 2023. Memory-augmented llm personalization with short-and long-term memory coordination. arXiv preprint arXiv:2309.11696. — [arXiv:2309.11696](https://arxiv.org/abs/2309.11696) [find](https://scholar.google.com/scholar?q=Kai+Zhang%2C+Fubang+Zhao%2C+Yangyang+Kang%2C+and+Xiaozhong+Liu%2E+2023%2E+Memory%2Daugmented+llm+personalization+with+short%2Dand+long%2Dterm+memory+coordination%2E+arXiv+preprint+arXiv%3A2309%2E11696%2E+%E2%80%94+%5BarXiv%3A2309%2E11696%5D)
40. Shaolei Zhang, Tian Yu, and Yang Feng. 2024b. [Truthx: Alleviating hallucinations by editing large language models in truthful space](https://md.research.elamur.ai/free/en/papers/zhang-2024-truthx-alleviating-hallucinations-editing-large/). arXiv preprint arXiv:2402.17811.
41. Zhehao Zhang, Ryan A Rossi, Branislav Kveton, Yijia Shao, Diyi Yang, Hamed Zamani, Franck Dernoncourt, Joe Barrow, Tong Yu, Sungchul Kim, et al. 2024c. Personalization of [large language models: A survey](https://md.research.elamur.ai/free/en/papers/minaee-2024-large-language-models-survey/). arXiv preprint arXiv:2411.00027.
42. Jianfeng Zhu, Ruoming Jin, and Karin G. Coifman. 2025. Investigating large language models in inferring personality traits from user conversations. ArXiv. [find](https://scholar.google.com/scholar?q=Jianfeng+Zhu%2C+Ruoming+Jin%2C+and+Karin+G%2E+Coifman%2E+2025%2E+Investigating+large+language+models+in+inferring+personality+traits+from+user+conversations%2E+ArXiv%2E)
43. 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. arXiv preprint arXiv:2406.02888. — [arXiv:2406.02888](https://arxiv.org/abs/2406.02888) [find](https://scholar.google.com/scholar?q=Yuchen+Zhuang%2C+Haotian+Sun%2C+Yue+Yu%2C+Rushi+Qiang%2C+Qifan+Wang%2C+Chao+Zhang%2C+and+Bo+Dai%2E+2024%2E+Hydra%3A+Model+factorization+framework+for+black%2Dbox+llm+personalization%2E+arXiv+preprint+arXiv%3A2406%2E02888%2E+%E2%80%94+%5Ba)
44. Algorithm 1 Personalized Generation with Style Steering [find](https://scholar.google.com/scholar?q=Algorithm+1+Personalized+Generation+with+Style+Steering)

*Updated 1 November 2026*