Latent Preference Modeling for Cross-Session Personalized Tool Calling
Fuente:
arXiv
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866917420929646592 |
|---|---|
| author | Yoon, Yejin Kim, Minseo Kim, Taeuk |
| author_facet | Yoon, Yejin Kim, Minseo Kim, Taeuk |
| contents | Users often omit essential details in their requests to LLM-based agents, resulting in under-specified inputs for tool use. This poses a fundamental challenge for tool-augmented agents, as API execution typically requires complete arguments, highlighting the need for personalized tool calling. To study this problem, we introduce MPT, a benchmark comprising 265 multi-session dialogues that cover three challenges: Preference Recall, Preference Induction, and Preference Transfer. We also propose PRefine, a test-time memory-augmented method that represents user preferences as evolving hypotheses. Through a generate--verify--refine loop, it extracts reusable constraints from history and improves tool-calling accuracy while using only 1.24% of the tokens required by full-history prompting. These results indicate that robust personalization in agentic systems depends on memory that captures the reasons behind user choices, not just the choices themselves. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_17886 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Latent Preference Modeling for Cross-Session Personalized Tool Calling Yoon, Yejin Kim, Minseo Kim, Taeuk Computation and Language Artificial Intelligence Users often omit essential details in their requests to LLM-based agents, resulting in under-specified inputs for tool use. This poses a fundamental challenge for tool-augmented agents, as API execution typically requires complete arguments, highlighting the need for personalized tool calling. To study this problem, we introduce MPT, a benchmark comprising 265 multi-session dialogues that cover three challenges: Preference Recall, Preference Induction, and Preference Transfer. We also propose PRefine, a test-time memory-augmented method that represents user preferences as evolving hypotheses. Through a generate--verify--refine loop, it extracts reusable constraints from history and improves tool-calling accuracy while using only 1.24% of the tokens required by full-history prompting. These results indicate that robust personalization in agentic systems depends on memory that captures the reasons behind user choices, not just the choices themselves. |
| title | Latent Preference Modeling for Cross-Session Personalized Tool Calling |
| topic | Computation and Language Artificial Intelligence |
| url | https://arxiv.org/abs/2604.17886 |