Mind2Report: A Cognitive Deep Research Agent for Expert-Level Commercial Report Synthesis
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_ | 1866915716497670144 |
|---|---|
| author | Cheng, Mingyue Wang, Daoyu Liu, Qi Yu, Shuo Tao, Xiaoyu Wang, Yuqian Chu, Chengzhong Duan, Yu Long, Mingkang Chen, Enhong |
| author_facet | Cheng, Mingyue Wang, Daoyu Liu, Qi Yu, Shuo Tao, Xiaoyu Wang, Yuqian Chu, Chengzhong Duan, Yu Long, Mingkang Chen, Enhong |
| contents | Synthesizing informative commercial reports from massive and noisy web sources is critical for high-stakes business decisions. Although current deep research agents achieve notable progress, their reports still remain limited in terms of quality, reliability, and coverage. In this work, we propose Mind2Report, a cognitive deep research agent that emulates the commercial analyst to synthesize expert-level reports. Specifically, it first probes fine-grained intent, then searches web sources and records distilled information on the fly, and subsequently iteratively synthesizes the report. We design Mind2Report as a training-free agentic workflow that augments general large language models (LLMs) with dynamic memory to support these long-form cognitive processes. To rigorously evaluate Mind2Report, we further construct QRC-Eval comprising 200 real-world commercial tasks and establish a holistic evaluation strategy to assess report quality, reliability, and coverage. Experiments demonstrate that Mind2Report outperforms leading baselines, including OpenAI and Gemini deep research agents. Although this is a preliminary study, we expect it to serve as a foundation for advancing the future design of commercial deep research agents. Our code and data are available at https://github.com/Melmaphother/Mind2Report. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_04879 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Mind2Report: A Cognitive Deep Research Agent for Expert-Level Commercial Report Synthesis Cheng, Mingyue Wang, Daoyu Liu, Qi Yu, Shuo Tao, Xiaoyu Wang, Yuqian Chu, Chengzhong Duan, Yu Long, Mingkang Chen, Enhong Computation and Language Synthesizing informative commercial reports from massive and noisy web sources is critical for high-stakes business decisions. Although current deep research agents achieve notable progress, their reports still remain limited in terms of quality, reliability, and coverage. In this work, we propose Mind2Report, a cognitive deep research agent that emulates the commercial analyst to synthesize expert-level reports. Specifically, it first probes fine-grained intent, then searches web sources and records distilled information on the fly, and subsequently iteratively synthesizes the report. We design Mind2Report as a training-free agentic workflow that augments general large language models (LLMs) with dynamic memory to support these long-form cognitive processes. To rigorously evaluate Mind2Report, we further construct QRC-Eval comprising 200 real-world commercial tasks and establish a holistic evaluation strategy to assess report quality, reliability, and coverage. Experiments demonstrate that Mind2Report outperforms leading baselines, including OpenAI and Gemini deep research agents. Although this is a preliminary study, we expect it to serve as a foundation for advancing the future design of commercial deep research agents. Our code and data are available at https://github.com/Melmaphother/Mind2Report. |
| title | Mind2Report: A Cognitive Deep Research Agent for Expert-Level Commercial Report Synthesis |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2601.04879 |