Elamur Research⌕
☰
home / papers / 2026 / D2PCM:A Multi-Turn Dialogue Dataset with Personalized Contextual Memory

D2PCM:A Multi-Turn Dialogue Dataset with Personalized Contextual Memory

Paper · 2026 · Zhe Yang, Yi Huang, Yaqin Chen et al.

Summary

Most memory-augmented dialogue systems retrieve from a single shared memory pool by query similarity and ignore how personality shapes what a person remembers and prefers. Drawing on the self-reference effect from cognitive psychology, the authors build D2PCM, a synthetic multi-turn dialogue benchmark where memory is personalized. User profiles come from the Big Five model, with each trait set to high or low. For every event, an LLM generates memory chunks in several personality styles. One style matches the user, and the others are perturbed variants. At each turn the pipeline generates a query, produces candidate memories, samples one memory, writes a response, and scores it with a reward model. The dataset therefore supports SFT as well as RL-style post-training. The authors evaluate a personalized memory-selection policy against RAG and test several post-training methods. Both approaches raise persona-aligned memory use and reward scores. The resource lets researchers study memory retrieval and personalization together.

Key points

  • Personas use five Big Five axes, each binarized, giving 32 archetypes. Each event has n = 5 memory personas: the user's own plus four perturbed variants. The user-matching chunk is sampled with probability 0.4, and the remaining 0.6 is split evenly across the others.
  • Each turn holds a query, a memory chunk, the chosen memory, and the assistant response. Candidate responses come with reward values, and evaluation uses LLM-judge win-rate plus memory and response rewards.
  • Compared with RAG, personalized memory selection on Llama-8b raised memory reward by 0.6412 (+15.9%) and response reward by 0.4130 (+16.5%). Qwen-8b gained +20.2% and +20.8% on the same two rewards.
  • Post-training on D2PCM improved win-rate by about 10–47%. SFT gave the smallest gains, and DPO gave the largest (47.10% on Llama-8b, 37.32% on Qwen-8b).
  • The best memory ratios were 0.8255 for GRPO on Llama-8b (+118.38%) and 0.8580 for DPO on Qwen-8b (+156.35%). The authors note that relying on LLM-as-a-judge is a limitation.

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

🎓Tsinghua UniversityBeijing · China415 works · 94,028 citations

 Fields, topics and concepts

📚 Shelves 4The 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 works

🧭 Topics 1Research subjects and languages the work is about.Dialogue systems · 654 works

Cite

Cite as Zhe Yang, Yi Huang, Yaqin Chen, Chunyang Gao, Jingyu Yao and Junlan Feng. 2026. D2PCM:A Multi-Turn Dialogue Dataset with Personalized Contextual Memory. 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.1870

@inproceedings{yang2026d2pcm,
  title = {{D2PCM:A Multi-Turn Dialogue Dataset with Personalized Contextual Memory}},
  author = {Zhe Yang, Yi Huang, Yaqin Chen, Chunyang Gao, Jingyu Yao and Junlan Feng},
  booktitle = {Findings of the Association for Computational Linguistics: ACL 2026},
  year = {2026},
  doi = {10.18653/v1/2026.findings-acl.1870},
  url = {https://doi.org/10.18653/v1/2026.findings-acl.1870}
}

References

35 references, 11 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Courtney E. Ackerman. 2020. Big five personality traits. Encyclopedia of Education and Information Technologies. find
  2. Akari Asai, Zeqiu Wu, Yizhong Wang, Avirup Sil, and Hannaneh Hajishirzi. 2024. Self-RAG: Learning to retrieve, generate, and critique through self-reflection. In The Twelfth International Conference on Learning Representations.
  3. Shulin Cao, Jiaxin Shi, Liangming Pan, Lun Yiu Nie, Yutong Xiang, Lei Hou, Juanzi Li, Bin He, and Hanwang Zhang. 2020. Kqa pro: A dataset with explicit compositional programs for complex question answering over knowledge base. In Annual Meeting of the Association for Computational Linguistics. find
  4. Yiming Du, Wenyu Huang, Danna Zheng, Zhaowei Wang, Sebastien Montella, Mirella Lapata, Kam-Fai Wong, and Jeff Z. Pan. 2025. Rethinking memory in ai: Taxonomy, operations, topics, and future directions. Preprint, arXiv:2505.00675. — [arXiv:2505.00675](https://arxiv.org/abs/2505.00675) find
  5. Yiming Du, Bingbing Wang, Yang He, Bin Liang, Baojun Wang, Zhongyang Li, Lin Gui, Jeff Z. Pan, Ruifeng Xu, and Kam-Fai Wong. 2026. Memguide: Intent-driven memory selection for goal-oriented multi-session llm agents. Proceedings of the AAAI Conference on Artificial Intelligence, 40(36):30584– 30592. find
  6. Yiming Du, Hongru Wang, Zhengyi Zhao, Bin Liang, Baojun Wang, Wanjun Zhong, Zezhong Wang, and Kam-Fai Wong. 2024. Perltqa: A personal long-term memory dataset for memory classification, retrieval, and synthesis in question answering. ArXiv, abs/2402.16288. — [arXiv:2402.16288](https://arxiv.org/abs/2402.16288) find
  7. Yunfan Gao, Yun Xiong, Xinyu Gao, Kangxiang Jia, Jinliu Pan, Yuxi Bi, Yi Dai, Jiawei Sun, Qianyu Guo, Meng Wang, and Haofen Wang. 2023. Retrieval-augmented generation for large language models: A survey. ArXiv, abs/2312.10997.
  8. Aaron Grattafiori, Abhimanyu Dubey, Abhinav Jauhri, Abhinav Pandey, Abhishek Kadian, Ahmad Al-Dahle, Aiesha Letman, Akhil Mathur, Alan Schelten, Alex Vaughan, Amy Yang, Angela Fan, Anirudh Goyal, Anthony Hartshorn, Aobo Yang, Archi Mitra, Archie Sravankumar, Artem Korenev, Arthur Hinsvark, and 542 o… — The Llama 3 Herd of Models
  9. Jiawei Gu, Xuhui Jiang, Zhichao Shi, Hexiang Tan, Xuehao Zhai, Chengjin Xu, Wei Li, Yinghan Shen, Shengjie Ma, Honghao Liu, Yuanzhuo Wang, and Jian Guo. 2024. A survey on llm-as-a-judge. ArXiv, abs/2411.15594. find
  10. Kostas Hatalis, Despina Christou, Joshua Myers, Steven Jones, Keith Lambert, Adam Amos-Binks, Zohreh Dannenhauer, and Dustin Dannenhauer. 2024. Memory matters: The need to improve long-term memory in llm-agents. Proceedings of the AAAI Symposium Series. find
  11. Matthew Thomas Jackson, Uljad Berdica, Jarek Luca Liesen, Shimon Whiteson, and Jakob Nicolaus Foerster. 2025. A clean slate for offline reinforcement learning. In The Thirty-ninth Annual Conference on Neural Information Processing Systems. find
  12. Jihyoung Jang, Taeyoung Kim, and Hyounghun Kim. 2024. Mixed-session conversation with egocentric memory. In Findings of the Association for Computational Linguistics: EMNLP 2024, pages 11786– 11815, Miami, Florida, USA. Association for Computational Linguistics. find
  13. Jiaming Ji, Boyuan Chen, Hantao Lou, Donghai Hong, Borong Zhang, Xuehai Pan, Tianyi Qiu, Juntao Dai, and Yaodong Yang. 2024. Aligner: Efficient alignment by learning to correct. In The Thirty-eighth Annual Conference on Neural Information Processing Systems.
  14. Eunwon Kim, Chanho Park, and Buru Chang. 2024. Share: Shared memory-aware open-domain long-term dialogue dataset constructed from movie script. In Annual Meeting of the Association for Computational Linguistics. find
  15. Patrick Lewis, Ethan Perez, Aleksandra Piktus, Fabio Petroni, Vladimir Karpukhin, Naman Goyal, Heinrich Küttler, Mike Lewis, Wen-tau Yih, Tim Rocktäschel, Sebastian Riedel, and Douwe Kiela. 2020. find
  16. Retrieval-augmented generation for knowledge-intensive nlp tasks. In Advances in Neural Information Processing Systems, volume 33, pages 9459– 9474. Curran Associates, Inc. 2023. Direct preference optimization: Your language model is secretly a reward model. In Thirty-seventh Conference on Neural In…
  17. Hao Li, Chenghao Yang, An Zhang, Yang Deng, Xiang Wang, and Tat-Seng Chua. 2025a. Hello again! LLM-powered personalized agent for long-term dialogue. In Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: Human Language Technolo…
  18. Timothy B. Rogers, Nicholas A. Kuiper, and W. S. Kirker. 1977. Self-reference and the encoding of personal information. Journal of personality and social psychology, 35 9:677–88. find
  19. Xintong Li, Jalend Bantupalli, Ria Dharmani, Yuwei Zhang, and Jingbo Shang. 2025b. Toward multi-session personalized conversation: A large-scale dataset and hierarchical tree framework for implicit reasoning. In Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing,… find
  20. Alex Troy Mallen, Akari Asai, Victor Zhong, Rajarshi Das, Hannaneh Hajishirzi, and Daniel Khashabi. 2022. When not to trust language models: Investigating effectiveness of parametric and non-parametric memories. In Annual Meeting of the Association for Computational Linguistics. find
  21. Youssef Mroueh, Nicolas Dupuis, Brian M. Belgodere, Apoorva Nitsure, Mattia Rigotti, Kristjan H. Greenewald, Jirí Navrátil, Jerret Ross, and Jesus Rios. 2025. Revisiting group relative policy optimization: Insights into on-policy and off-policy training. ArXiv, abs/2505.22257. — [arXiv:2505.22257](https://arxiv.org/abs/2505.22257) find
  22. Long Ouyang, Jeff Wu, Xu Jiang, Diogo Almeida, Carroll L. Wainwright, Pamela Mishkin, Chong Zhang, Sandhini Agarwal, Katarina Slama, Alex Ray, John Schulman, Jacob Hilton, Fraser Kelton, Luke Miller, Maddie Simens, Amanda Askell, Peter Welinder, Paul Christiano, Jan Leike, and Ryan Lowe. 2022. Train…
  23. Zhuoshi Pan, Qianhui Wu, Huiqiang Jiang, Xufang Luo, Hao Cheng, Dongsheng Li, Yuqing Yang, Chin-Yew Lin, H. Vicky Zhao, Lili Qiu, and Jianfeng Gao. 2025. Secom: On memory construction and retrieval for personalized conversational agents. In The Thirteenth International Conference on Learning Represe… find
  24. Rafael Figueiredo Prudencio, Marcos R.O.A. Maximo, and Esther Luna Colombini. 2022. A survey on offline reinforcement learning: Taxonomy, review, and open problems. IEEE Transactions on Neural Networks and Learning Systems, 35:10237–10257. find
  25. Rafael Rafailov, Archit Sharma, Eric Mitchell, Christopher D Manning, Stefano Ermon, and Chelsea Finn. find
  26. Zhihong Shao, Peiyi Wang, Qihao Zhu, Runxin Xu, Junxiao Song, Xiao Bi, Haowei Zhang, Mingchuan Zhang, Y. K. Li, Y. Wu, and Daya Guo. 2024. Deepseekmath: Pushing the limits of mathematical reasoning in open language models. Preprint, arXiv:2402.03300.
  27. Alon Talmor and Jonathan Berant. 2018. The web as a knowledge-base for answering complex questions. In Proceedings of the 2018 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long Papers), pages 641–651.
  28. Huaijie Wang, Shibo Hao, Hanze Dong, Shenao Zhang, Yilin Bao, Ziran Yang, and Yi Wu. 2025a. Offline reinforcement learning for LLM multi-step reasoning. In Findings of the Association for Computational Linguistics: ACL 2025, pages 8881–8893, Vienna, Austria. Association for Computational Linguistics… find
  29. Yilei Wang, Jiabao Zhao, Deniz S. Ones, Liang He, and Xin Xu. 2025b. Evaluating the ability of large language models to emulate personality. Scientific Reports, 15. find
  30. Zilong Wang, Zifeng Wang, Long Le, Steven Zheng, Swaroop Mishra, Vincent Perot, Yuwei Zhang, Anush Mattapalli, Ankur Taly, Jingbo Shang, Chen-Yu Lee, and Tomas Pfister. 2025c. Speculative RAG: Enhancing retrieval augmented generation through drafting. In The Thirteenth International Conference on Le… find
  31. Charlotte Grosse Wiesmann, Katrin Rothmaler, Esra Hasan, Kathrine Habdank, Chen Yang, Emanuela Yeung, and Victoria Southgate. 2025. The self-reference memory bias is preceded by an other-reference bias in infancy. Nature Communications, 16(1):1–8. find
  32. Wujiang Xu, Zujie Liang, Kai Mei, Hang Gao, Juntao Tan, and Yongfeng Zhang. 2025. A-mem: Agentic memory for LLM agents. In The Thirty-ninth Annual Conference on Neural Information Processing Systems.
  33. An Yang, Anfeng Li, Baosong Yang, Beichen Zhang, Binyuan Hui, Bo Zheng, Bowen Yu, Chang Gao, Chengen Huang, Chenxu Lv, Chujie Zheng, Dayiheng Liu, Fan Zhou, Fei Huang, Feng Hu, Hao Ge, Haoran Wei, Huan Lin, Jialong Tang, and 41 others. 2025. Qwen3 technical report. Preprint, arXiv:2505.09388.
  34. Ruifeng Yuan, Shichao Sun, Yongqi Li, Zili Wang, Ziqiang Cao, and Wenjie Li. 2024. Personalized large language model assistant with evolving conditional memory. Preprint, arXiv:2312.17257. find
  35. Wanjun Zhong, Lianghong Guo, Qiqi Gao, He Ye, and Yanlin Wang. 2024. Memorybank: enhancing large language models with long-term memory. In Proceedings of the Thirty-Eighth AAAI Conference on Artificial Intelligence and Thirty-Sixth Conference on Innovative Applications of Artificial Intelligence and… find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/yang-2026-d2pcm-multi-turn-dialogue-dataset/