SynPlanResearch-R1: Encouraging Tool Exploration for Deep Research with Synthetic Plans
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_ | 1866915845481955328 |
|---|---|
| author | Zeng, Hansi Li, Zoey Gao, Yifan Zhang, Chenwei Pan, Xiaoman Yang, Tao Mo, Fengran Lin, Jiacheng Li, Xian Shang, Jingbo |
| author_facet | Zeng, Hansi Li, Zoey Gao, Yifan Zhang, Chenwei Pan, Xiaoman Yang, Tao Mo, Fengran Lin, Jiacheng Li, Xian Shang, Jingbo |
| contents | Research Agents enable models to gather information from the web using tools to answer user queries, requiring them to dynamically interleave internal reasoning with tool use. While such capabilities can in principle be learned via reinforcement learning with verifiable rewards (RLVR), we observe that agents often exhibit poor exploration behaviors, including premature termination and biased tool usage. As a result, RLVR alone yields limited improvements. We propose SynPlanResearch-R1, a framework that synthesizes tool-use trajectories that encourage deeper exploration to shape exploration during cold-start supervised fine-tuning, providing a strong initialization for subsequent RL. Across seven multi-hop and open-web benchmarks, \framework improves performance by up to 6.0% on Qwen3-8B and 5.8% on Qwen3-4B backbones respectively compared to SOTA baselines. Further analyses of tool-use patterns and training dynamics compared to baselines shed light on the factors underlying these gains. Our code is publicly available at https://github.com/HansiZeng/syn-plan-research. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_07853 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | SynPlanResearch-R1: Encouraging Tool Exploration for Deep Research with Synthetic Plans Zeng, Hansi Li, Zoey Gao, Yifan Zhang, Chenwei Pan, Xiaoman Yang, Tao Mo, Fengran Lin, Jiacheng Li, Xian Shang, Jingbo Artificial Intelligence Computation and Language Information Retrieval Research Agents enable models to gather information from the web using tools to answer user queries, requiring them to dynamically interleave internal reasoning with tool use. While such capabilities can in principle be learned via reinforcement learning with verifiable rewards (RLVR), we observe that agents often exhibit poor exploration behaviors, including premature termination and biased tool usage. As a result, RLVR alone yields limited improvements. We propose SynPlanResearch-R1, a framework that synthesizes tool-use trajectories that encourage deeper exploration to shape exploration during cold-start supervised fine-tuning, providing a strong initialization for subsequent RL. Across seven multi-hop and open-web benchmarks, \framework improves performance by up to 6.0% on Qwen3-8B and 5.8% on Qwen3-4B backbones respectively compared to SOTA baselines. Further analyses of tool-use patterns and training dynamics compared to baselines shed light on the factors underlying these gains. Our code is publicly available at https://github.com/HansiZeng/syn-plan-research. |
| title | SynPlanResearch-R1: Encouraging Tool Exploration for Deep Research with Synthetic Plans |
| topic | Artificial Intelligence Computation and Language Information Retrieval |
| url | https://arxiv.org/abs/2603.07853 |