Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2603.28197 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866908920665079808 |
|---|---|
| author | Zhang, Yujie Yuan, Weikang Jiang, Zhuoren Yan, Pengwei |
| author_facet | Zhang, Yujie Yuan, Weikang Jiang, Zhuoren Yan, Pengwei |
| contents | Pluralistic alignment is essential for adapting large language models (LLMs) to the diverse preferences of individuals and minority groups. However, existing approaches often mix stable personal traits with episode-specific factors, limiting their ability to generalize across episodes. To address this challenge, we introduce EpiPersona, a framework for explicit persona-episode coupling. EpiPersona first projects noisy preference feedback into a low-dimensional persona space, where similar personas are aggregated into shared discrete codes. This process separates enduring personal characteristics from situational signals without relying on predefined preference dimensions. The inferred persona representation is then coupled with the current episode, enabling episode-aware preference prediction. Extensive experiments show that EpiPersona consistently outperforms the baselines. It achieves notable performance gains in hard episodic-shift scenarios, while remaining effective with sparse preference data. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_28197 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | EpiPersona: Persona Projection and Episode Coupling for Pluralistic Preference Modeling Zhang, Yujie Yuan, Weikang Jiang, Zhuoren Yan, Pengwei Artificial Intelligence Pluralistic alignment is essential for adapting large language models (LLMs) to the diverse preferences of individuals and minority groups. However, existing approaches often mix stable personal traits with episode-specific factors, limiting their ability to generalize across episodes. To address this challenge, we introduce EpiPersona, a framework for explicit persona-episode coupling. EpiPersona first projects noisy preference feedback into a low-dimensional persona space, where similar personas are aggregated into shared discrete codes. This process separates enduring personal characteristics from situational signals without relying on predefined preference dimensions. The inferred persona representation is then coupled with the current episode, enabling episode-aware preference prediction. Extensive experiments show that EpiPersona consistently outperforms the baselines. It achieves notable performance gains in hard episodic-shift scenarios, while remaining effective with sparse preference data. |
| title | EpiPersona: Persona Projection and Episode Coupling for Pluralistic Preference Modeling |
| topic | Artificial Intelligence |
| url | https://arxiv.org/abs/2603.28197 |