AI Paper Atlas

Research map / Multi-agent systems and games / LLM agents

DeepWriter: A Multi-Agent Collaboration Framework for Information-rich Ultra-long Book Writing

Ming Wang, Minghao Hu, Xiuli Kang, Li He, Yu Tian, Chunming Liu, Han Shi, Zhunchen Luo, Wei Luo, Guotong Geng

AAAI 2026 · Multi-agent systems and games › LLM agents · approach: Agent data synthesis and behavior studies · 1 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Proposes DeepWriter, a multi-agent collaboration framework for generating information-rich, ultra-long books (100k+ words).

Problem

Existing long-text generation methods cannot produce ultra-long, book-scale content and lack mechanisms to integrate rich factual information.

Approach

A multi-agent collaboration framework for book writing that targets both book-scale length and integration of factual information into the generated text.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add OutlineBot planning and retrieval-augmented WriteBot (ablation)Direct Gen · DeepWriter-BenchBookScore: 31.93 -> 58.35Improved
add TriviaBot (ablation)RAG Gen · DeepWriter-BenchBookScore: 58.35 -> 67.54Improved
add ImageBot (ablation)DeepWriter Trivia · DeepWriter-BenchBookScore: 67.54 -> 80.92Improved
replace Co-STORM with full DeepWriterCo-STORM · DeepWriter-BenchBookScore: 47.23 -> 80.92Improved
use OutlineBot expert-panel deliberation instead of Co-STORMOutline generation · DeepWriter-Benchheading soft recall: 79.65 -> 89.37Improved
use OutlineBot instead of Direct Gen outlineOutline generation · DeepWriter-Benchheading entity recall: 31.26 -> 62.12Improved
use WriteBot instead of Co-STORMMain text · DeepWriter-BenchROUGE-1: 45.32 -> 60.34Improved
use WriteBot instead of Co-STORMCitation verifiability · DeepWriter-Benchcitation recall: 65.73 -> 77.35Improved
add context-aware retrieval for trivia (TriviaBot vs Trivia RAG) (ablation)Trivia generation · DeepWriter-Benchfactual accuracy: 4.03 -> 4.63Improved
add retrieval to trivia generation (Trivia RAG vs Trivia Direct) (ablation)Trivia generation · DeepWriter-Benchfactual accuracy: 3.41 -> 4.03Improved
use ImageBot instead of Image FuzzyCrawl (ablation)Image insertion · DeepWriter-Benchsemantic consistency: 3.45 -> 4.32Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy