O-Researcher: An Open Ended Deep Research Model via Multi-Agent Distillation and Agentic RL
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_ | 1866912806730727424 |
|---|---|
| author | Yao, Yi Zhu, He Wang, Piaohong Ren, Jincheng Yang, Xinlong Chen, Qianben Li, Xiaowan Shi, Dingfeng Li, Jiaxian Wang, Qiexiang Wang, Sinuo Liu, Xinpeng Wu, Jiaqi Liu, Minghao Zhou, Wangchunshu |
| author_facet | Yao, Yi Zhu, He Wang, Piaohong Ren, Jincheng Yang, Xinlong Chen, Qianben Li, Xiaowan Shi, Dingfeng Li, Jiaxian Wang, Qiexiang Wang, Sinuo Liu, Xinpeng Wu, Jiaqi Liu, Minghao Zhou, Wangchunshu |
| contents | The performance gap between closed-source and open-source large language models (LLMs) is largely attributed to disparities in access to high-quality training data. To bridge this gap, we introduce a novel framework for the automated synthesis of sophisticated, research-grade instructional data. Our approach centers on a multi-agent workflow where collaborative AI agents simulate complex tool-integrated reasoning to generate diverse and high-fidelity data end-to-end. Leveraging this synthesized data, we develop a two-stage training strategy that integrates supervised fine-tuning with a novel reinforcement learning method, designed to maximize model alignment and capability. Extensive experiments demonstrate that our framework empowers open-source models across multiple scales, enabling them to achieve new state-of-the-art performance on the major deep research benchmark. This work provides a scalable and effective pathway for advancing open-source LLMs without relying on proprietary data or models. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_03743 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | O-Researcher: An Open Ended Deep Research Model via Multi-Agent Distillation and Agentic RL Yao, Yi Zhu, He Wang, Piaohong Ren, Jincheng Yang, Xinlong Chen, Qianben Li, Xiaowan Shi, Dingfeng Li, Jiaxian Wang, Qiexiang Wang, Sinuo Liu, Xinpeng Wu, Jiaqi Liu, Minghao Zhou, Wangchunshu Computation and Language Artificial Intelligence The performance gap between closed-source and open-source large language models (LLMs) is largely attributed to disparities in access to high-quality training data. To bridge this gap, we introduce a novel framework for the automated synthesis of sophisticated, research-grade instructional data. Our approach centers on a multi-agent workflow where collaborative AI agents simulate complex tool-integrated reasoning to generate diverse and high-fidelity data end-to-end. Leveraging this synthesized data, we develop a two-stage training strategy that integrates supervised fine-tuning with a novel reinforcement learning method, designed to maximize model alignment and capability. Extensive experiments demonstrate that our framework empowers open-source models across multiple scales, enabling them to achieve new state-of-the-art performance on the major deep research benchmark. This work provides a scalable and effective pathway for advancing open-source LLMs without relying on proprietary data or models. |
| title | O-Researcher: An Open Ended Deep Research Model via Multi-Agent Distillation and Agentic RL |
| topic | Computation and Language Artificial Intelligence |
| url | https://arxiv.org/abs/2601.03743 |