DynSess: Dynamic Session-Level Evaluation and Optimization Framework for Role-Playing Agents
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2026
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866910268961849344 |
|---|---|
| author | Zhang, Rongsheng Tang, Jiji Ren, Junnan Bao, Zuyi Chen, Weijie Hu, Ruofan Zhao, Zhou Lv, Tangjie Zhang, Yan |
| author_facet | Zhang, Rongsheng Tang, Jiji Ren, Junnan Bao, Zuyi Chen, Weijie Hu, Ruofan Zhao, Zhou Lv, Tangjie Zhang, Yan |
| contents | Role-playing with large language models is fundamentally a session-level task, requiring agents to sustain character identity and interaction quality across extended multi-turn conversations. Yet existing evaluation and optimization methods remain largely turn-level, failing to capture long-horizon quality. We propose DynSess, a unified session-level framework for role-playing agents. DynSess-Eval scores complete dialogue sessions via rubrics targeting long-horizon behaviors. Leveraging its session-level rewards, we construct high-quality training trajectories through multi-turn lookahead search and train DynSess-Character with two complementary variants: DSPO (off-policy) and GSRPO (on-policy). Experiments show that DynSess-Eval aligns with human judgments substantially better than prior evaluators, and blind human evaluation further shows that DynSess-Character matches the strongest character model despite using substantially fewer parameters, while maintaining strong role consistency and interactive ability. Our dataset and code will be released to facilitate future research. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_29256 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | DynSess: Dynamic Session-Level Evaluation and Optimization Framework for Role-Playing Agents Zhang, Rongsheng Tang, Jiji Ren, Junnan Bao, Zuyi Chen, Weijie Hu, Ruofan Zhao, Zhou Lv, Tangjie Zhang, Yan Computation and Language Artificial Intelligence Role-playing with large language models is fundamentally a session-level task, requiring agents to sustain character identity and interaction quality across extended multi-turn conversations. Yet existing evaluation and optimization methods remain largely turn-level, failing to capture long-horizon quality. We propose DynSess, a unified session-level framework for role-playing agents. DynSess-Eval scores complete dialogue sessions via rubrics targeting long-horizon behaviors. Leveraging its session-level rewards, we construct high-quality training trajectories through multi-turn lookahead search and train DynSess-Character with two complementary variants: DSPO (off-policy) and GSRPO (on-policy). Experiments show that DynSess-Eval aligns with human judgments substantially better than prior evaluators, and blind human evaluation further shows that DynSess-Character matches the strongest character model despite using substantially fewer parameters, while maintaining strong role consistency and interactive ability. Our dataset and code will be released to facilitate future research. |
| title | DynSess: Dynamic Session-Level Evaluation and Optimization Framework for Role-Playing Agents |
| topic | Computation and Language Artificial Intelligence |
| url | https://arxiv.org/abs/2605.29256 |