@inproceedings{liu-etal-2026-deepdiver,
title = "{D}eep{D}iver-{DR}: Empowering {LLM}s for Deep Research via Specialized Multi-Agent Collaboration",
author = "Liu, Zengyan and
Wu, Han and
Xu, Haonan and
Shi, Wenxuan and
Wang, Jingyi and
Tan, Haochen and
Gao, Jiahui and
Bai, Haoli and
Li, Xiaoguang and
Song, Linqi",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {EMNLP} 2026",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.findings-emnlp.1154/",
pages = "24421--24435",
ISBN = "979-8-89176-462-0",
abstract = "Deep Research (DR) systems iteratively plan, retrieve evidence, and synthesize long-form reports in response to complex, open-ended queries. Existing approaches suffer from context overload, role entanglement, and error propagation over long horizons, while the most capable DR systems remain locked behind massive proprietary models. We present DeepDiver-DR (DD-DR), a multi-agent framework that decomposes the DR process into three specialized roles{---}planner, information seeker, and writer{---}coordinated through a persistent, file-based shared workspace. Each role is further equipped with fine-grained, DR-specific tools. To support rigorous evaluation, we introduce RubricWrite-Bench (RW-Bench), a benchmark of 100 rubric-based writing queries with a two-phase assessment of global report quality and point-level rubric fulfillment. Experiments show that DD-DR (30B) establishes a new state of the art among models at the same parameter scale on the DeepResearch Bench, and achieves performance competitive with leading proprietary DR systems on RW-Bench. Ablations further confirm the contribution of the architectural components."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="liu-etal-2026-deepdiver">
<titleInfo>
<title>DeepDiver-DR: Empowering LLMs for Deep Research via Specialized Multi-Agent Collaboration</title>
</titleInfo>
<name type="personal">
<namePart type="given">Zengyan</namePart>
<namePart type="family">Liu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Han</namePart>
<namePart type="family">Wu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Haonan</namePart>
<namePart type="family">Xu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Wenxuan</namePart>
<namePart type="family">Shi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jingyi</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Haochen</namePart>
<namePart type="family">Tan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jiahui</namePart>
<namePart type="family">Gao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Haoli</namePart>
<namePart type="family">Bai</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xiaoguang</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Linqi</namePart>
<namePart type="family">Song</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-10</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: EMNLP 2026</title>
</titleInfo>
<name type="personal">
<namePart type="given">Hung-Yi</namePart>
<namePart type="family">Lee</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sunipa</namePart>
<namePart type="family">Dev</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jessy</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Desmond</namePart>
<namePart type="family">Elliott</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Budapest, Hungary</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-462-0</identifier>
</relatedItem>
<abstract>Deep Research (DR) systems iteratively plan, retrieve evidence, and synthesize long-form reports in response to complex, open-ended queries. Existing approaches suffer from context overload, role entanglement, and error propagation over long horizons, while the most capable DR systems remain locked behind massive proprietary models. We present DeepDiver-DR (DD-DR), a multi-agent framework that decomposes the DR process into three specialized roles—planner, information seeker, and writer—coordinated through a persistent, file-based shared workspace. Each role is further equipped with fine-grained, DR-specific tools. To support rigorous evaluation, we introduce RubricWrite-Bench (RW-Bench), a benchmark of 100 rubric-based writing queries with a two-phase assessment of global report quality and point-level rubric fulfillment. Experiments show that DD-DR (30B) establishes a new state of the art among models at the same parameter scale on the DeepResearch Bench, and achieves performance competitive with leading proprietary DR systems on RW-Bench. Ablations further confirm the contribution of the architectural components.</abstract>
<identifier type="citekey">liu-etal-2026-deepdiver</identifier>
<location>
<url>https://aclanthology.org/2026.findings-emnlp.1154/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>24421</start>
<end>24435</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T DeepDiver-DR: Empowering LLMs for Deep Research via Specialized Multi-Agent Collaboration
%A Liu, Zengyan
%A Wu, Han
%A Xu, Haonan
%A Shi, Wenxuan
%A Wang, Jingyi
%A Tan, Haochen
%A Gao, Jiahui
%A Bai, Haoli
%A Li, Xiaoguang
%A Song, Linqi
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Findings of the Association for Computational Linguistics: EMNLP 2026
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-462-0
%F liu-etal-2026-deepdiver
%X Deep Research (DR) systems iteratively plan, retrieve evidence, and synthesize long-form reports in response to complex, open-ended queries. Existing approaches suffer from context overload, role entanglement, and error propagation over long horizons, while the most capable DR systems remain locked behind massive proprietary models. We present DeepDiver-DR (DD-DR), a multi-agent framework that decomposes the DR process into three specialized roles—planner, information seeker, and writer—coordinated through a persistent, file-based shared workspace. Each role is further equipped with fine-grained, DR-specific tools. To support rigorous evaluation, we introduce RubricWrite-Bench (RW-Bench), a benchmark of 100 rubric-based writing queries with a two-phase assessment of global report quality and point-level rubric fulfillment. Experiments show that DD-DR (30B) establishes a new state of the art among models at the same parameter scale on the DeepResearch Bench, and achieves performance competitive with leading proprietary DR systems on RW-Bench. Ablations further confirm the contribution of the architectural components.
%U https://aclanthology.org/2026.findings-emnlp.1154/
%P 24421-24435
Markdown (Informal)
[DeepDiver-DR: Empowering LLMs for Deep Research via Specialized Multi-Agent Collaboration](https://aclanthology.org/2026.findings-emnlp.1154/) (Liu et al., Findings 2026)
ACL
- Zengyan Liu, Han Wu, Haonan Xu, Wenxuan Shi, Jingyi Wang, Haochen Tan, Jiahui Gao, Haoli Bai, Xiaoguang Li, and Linqi Song. 2026. DeepDiver-DR: Empowering LLMs for Deep Research via Specialized Multi-Agent Collaboration. In Findings of the Association for Computational Linguistics: EMNLP 2026, pages 24421–24435, Budapest, Hungary. Association for Computational Linguistics.