Saved in:
Bibliographic Details
Main Authors: Tian, Kuo, Sun, Pengfei, Wu, Zhen, Ding, Junran, Dai, Xinyu
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2604.17072
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914494112858112
author Tian, Kuo
Sun, Pengfei
Wu, Zhen
Ding, Junran
Dai, Xinyu
author_facet Tian, Kuo
Sun, Pengfei
Wu, Zhen
Ding, Junran
Dai, Xinyu
contents The autonomous synthesis of deep research reports represents a critical frontier for Large Language Models (LLMs), demanding sophisticated information orchestration and non-linear narrative logic. Current approaches rely on rigid predefined linear workflows, which cause error accumulation, preclude global restructuring from subsequent insights, and ultimately limit in-depth multimodal fusion and report quality. We propose CogGen, a Cognitively inspired recursive framework for deep research report Generation. Leveraging a Hierarchical Recursive Architecture to simulate cognitive writing, CogGen enables flexible planning and global restructuring. To extend this recursivity to multimodal content, we introduce Abstract Visual Representation (AVR): a concise intent-driven language that iteratively refines visual-text layouts without pixel-level regeneration overhead. We further present CLEF, a Cognitive Load Evaluation Framework, and curate a new benchmark from Our World in Data (OWID). Extensive experiments show CogGen achieves state-of-the-art results among open-source systems, generating reports comparable to professional analysts' outputs and surpassing Gemini Deep Research. Our code and dataset are available at https://github.com/NJUNLP/CogGen.
format Preprint
id arxiv_https___arxiv_org_abs_2604_17072
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle CogGen: A Cognitively Inspired Recursive Framework for Deep Research Report Generation
Tian, Kuo
Sun, Pengfei
Wu, Zhen
Ding, Junran
Dai, Xinyu
Multiagent Systems
The autonomous synthesis of deep research reports represents a critical frontier for Large Language Models (LLMs), demanding sophisticated information orchestration and non-linear narrative logic. Current approaches rely on rigid predefined linear workflows, which cause error accumulation, preclude global restructuring from subsequent insights, and ultimately limit in-depth multimodal fusion and report quality. We propose CogGen, a Cognitively inspired recursive framework for deep research report Generation. Leveraging a Hierarchical Recursive Architecture to simulate cognitive writing, CogGen enables flexible planning and global restructuring. To extend this recursivity to multimodal content, we introduce Abstract Visual Representation (AVR): a concise intent-driven language that iteratively refines visual-text layouts without pixel-level regeneration overhead. We further present CLEF, a Cognitive Load Evaluation Framework, and curate a new benchmark from Our World in Data (OWID). Extensive experiments show CogGen achieves state-of-the-art results among open-source systems, generating reports comparable to professional analysts' outputs and surpassing Gemini Deep Research. Our code and dataset are available at https://github.com/NJUNLP/CogGen.
title CogGen: A Cognitively Inspired Recursive Framework for Deep Research Report Generation
topic Multiagent Systems
url https://arxiv.org/abs/2604.17072