Saved in:
Bibliographic Details
Main Authors: Han, Rujun, Chen, Yanfei, CuiZhu, Zoey, Miculicich, Lesly, Sun, Guan, Bi, Yuanjun, Wen, Weiming, Wan, Hui, Wen, Chunfeng, Maître, Solène, Lee, George, Tirumalashetty, Vishy, Xue, Emily, Zhang, Zizhao, Haykal, Salem, Gokturk, Burak, Pfister, Tomas, Lee, Chen-Yu
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2507.16075
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915404103811072
author Han, Rujun
Chen, Yanfei
CuiZhu, Zoey
Miculicich, Lesly
Sun, Guan
Bi, Yuanjun
Wen, Weiming
Wan, Hui
Wen, Chunfeng
Maître, Solène
Lee, George
Tirumalashetty, Vishy
Xue, Emily
Zhang, Zizhao
Haykal, Salem
Gokturk, Burak
Pfister, Tomas
Lee, Chen-Yu
author_facet Han, Rujun
Chen, Yanfei
CuiZhu, Zoey
Miculicich, Lesly
Sun, Guan
Bi, Yuanjun
Wen, Weiming
Wan, Hui
Wen, Chunfeng
Maître, Solène
Lee, George
Tirumalashetty, Vishy
Xue, Emily
Zhang, Zizhao
Haykal, Salem
Gokturk, Burak
Pfister, Tomas
Lee, Chen-Yu
contents Deep research agents, powered by Large Language Models (LLMs), are rapidly advancing; yet, their performance often plateaus when generating complex, long-form research reports using generic test-time scaling algorithms. Drawing inspiration from the iterative nature of human research, which involves cycles of searching, reasoning, and revision, we propose the Test-Time Diffusion Deep Researcher (TTD-DR). This novel framework conceptualizes research report generation as a diffusion process. TTD-DR initiates this process with a preliminary draft, an updatable skeleton that serves as an evolving foundation to guide the research direction. The draft is then iteratively refined through a "denoising" process, which is dynamically informed by a retrieval mechanism that incorporates external information at each step. The core process is further enhanced by a self-evolutionary algorithm applied to each component of the agentic workflow, ensuring the generation of high-quality context for the diffusion process. This draft-centric design makes the report writing process more timely and coherent while reducing information loss during the iterative search process. We demonstrate that our TTD-DR achieves state-of-the-art results on a wide array of benchmarks that require intensive search and multi-hop reasoning, significantly outperforming existing deep research agents.
format Preprint
id arxiv_https___arxiv_org_abs_2507_16075
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Deep Researcher with Test-Time Diffusion
Han, Rujun
Chen, Yanfei
CuiZhu, Zoey
Miculicich, Lesly
Sun, Guan
Bi, Yuanjun
Wen, Weiming
Wan, Hui
Wen, Chunfeng
Maître, Solène
Lee, George
Tirumalashetty, Vishy
Xue, Emily
Zhang, Zizhao
Haykal, Salem
Gokturk, Burak
Pfister, Tomas
Lee, Chen-Yu
Computation and Language
Deep research agents, powered by Large Language Models (LLMs), are rapidly advancing; yet, their performance often plateaus when generating complex, long-form research reports using generic test-time scaling algorithms. Drawing inspiration from the iterative nature of human research, which involves cycles of searching, reasoning, and revision, we propose the Test-Time Diffusion Deep Researcher (TTD-DR). This novel framework conceptualizes research report generation as a diffusion process. TTD-DR initiates this process with a preliminary draft, an updatable skeleton that serves as an evolving foundation to guide the research direction. The draft is then iteratively refined through a "denoising" process, which is dynamically informed by a retrieval mechanism that incorporates external information at each step. The core process is further enhanced by a self-evolutionary algorithm applied to each component of the agentic workflow, ensuring the generation of high-quality context for the diffusion process. This draft-centric design makes the report writing process more timely and coherent while reducing information loss during the iterative search process. We demonstrate that our TTD-DR achieves state-of-the-art results on a wide array of benchmarks that require intensive search and multi-hop reasoning, significantly outperforming existing deep research agents.
title Deep Researcher with Test-Time Diffusion
topic Computation and Language
url https://arxiv.org/abs/2507.16075