Character-R1: Enhancing Role-Aware Reasoning in Role-Playing Agents via RLVR
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2026
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866917189884313600 |
|---|---|
| author | Tang, Yihong Chen, Kehai Bai, Xuefeng Wang, Benyou Liu, Zeming Wang, Haifeng Zhang, Min |
| author_facet | Tang, Yihong Chen, Kehai Bai, Xuefeng Wang, Benyou Liu, Zeming Wang, Haifeng Zhang, Min |
| contents | Current role-playing agents (RPAs) are typically constructed by imitating surface-level behaviors, but this approach lacks internal cognitive consistency, often causing out-of-character errors in complex situations. To address this, we propose Character-R1, a framework designed to provide comprehensive verifiable reward signals for effective role-aware reasoning, which are missing in recent studies. Specifically, our framework comprises three core designs: (1) Cognitive Focus Reward, which enforces explicit label-based analysis of 10 character elements (e.g., worldview) to structure internal cognition; (2) Reference-Guided Reward, which utilizes overlap-based metrics with reference responses as optimization anchors to enhance exploration and performance; and (3) Character-Conditioned Reward Normalization, which adjusts reward distributions based on character categories to ensure robust optimization across heterogeneous roles. Extensive experiments demonstrate that Character-R1 significantly outperforms existing methods in knowledge, memory and others. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_04611 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Character-R1: Enhancing Role-Aware Reasoning in Role-Playing Agents via RLVR Tang, Yihong Chen, Kehai Bai, Xuefeng Wang, Benyou Liu, Zeming Wang, Haifeng Zhang, Min Computation and Language Current role-playing agents (RPAs) are typically constructed by imitating surface-level behaviors, but this approach lacks internal cognitive consistency, often causing out-of-character errors in complex situations. To address this, we propose Character-R1, a framework designed to provide comprehensive verifiable reward signals for effective role-aware reasoning, which are missing in recent studies. Specifically, our framework comprises three core designs: (1) Cognitive Focus Reward, which enforces explicit label-based analysis of 10 character elements (e.g., worldview) to structure internal cognition; (2) Reference-Guided Reward, which utilizes overlap-based metrics with reference responses as optimization anchors to enhance exploration and performance; and (3) Character-Conditioned Reward Normalization, which adjusts reward distributions based on character categories to ensure robust optimization across heterogeneous roles. Extensive experiments demonstrate that Character-R1 significantly outperforms existing methods in knowledge, memory and others. |
| title | Character-R1: Enhancing Role-Aware Reasoning in Role-Playing Agents via RLVR |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2601.04611 |