Struct-Searcher: Agentic Structural Thinking Advances Multimodal Deep Information Seeking

Hugging Face Daily Papers Papers
multimodal information-seeking agentic-workflow belief-revision deep-research vision-language

Summary

Struct-Searcher introduces a belief revision theory-based structural agentic workflow for multimodal deep information seeking, achieving significant accuracy improvements over existing vision-language models and deep research agents.

Deep research agents have attracted increasing attention for their ability to collect large-scale online information to acquire target knowledge, with recent efforts shifting from purely text-based information seeking to multimodal settings. However, existing agentic workflows are largely aligned with evidence accumulation models, which linearly aggregate evidence and lack principled mechanisms for handling contradictory information across heterogeneous modalities. Towards this end, we propose Struct-Searcher, a structural agentic workflow grounded in belief revision theory that explicitly maintains an evolving multimodal structural graph throughout the reasoning process, enabling effective conflict-aware multimodal deep information seeking. Extensive experiments across multiple benchmark datasets and backbone models demonstrate that Struct-Searcher is (1) plug-and-play and model-agnostic, yielding an average relative accuracy improvement of 17.2% on BrowseComp-VL across five different backbones. (2) top-performing, consistently outperforming state-of-the-art vision-language models (VLMs) and deep research agents, with relative accuracy improvements of 3.7% on MM-BrowseComp, 1.5% on HLE-VL, and 0.7% on BrowseComp-VL over the second-best competing approach.
Original Article
View Cached Full Text

Cached at: 06/10/26, 05:45 AM

Paper page - Struct-Searcher: Agentic Structural Thinking Advances Multimodal Deep Information Seeking

Source: https://huggingface.co/papers/2606.07689

Abstract

Struct-Searcher introduces a belief revision theory-based structural agentic workflow for multimodal information seeking that improves accuracy over existing vision-language models and deep research agents.

Deep research agentshave attracted increasing attention for their ability to collect large-scale online information to acquire target knowledge, with recent efforts shifting from purely text-based information seeking to multimodal settings. However, existing agentic workflows are largely aligned withevidence accumulation models, which linearly aggregate evidence and lack principled mechanisms for handling contradictory information across heterogeneous modalities. Towards this end, we propose Struct-Searcher, astructural agentic workflowgrounded inbelief revision theorythat explicitly maintains an evolvingmultimodal structural graphthroughout the reasoning process, enabling effective conflict-aware multimodal deep information seeking. Extensive experiments across multiple benchmark datasets and backbone models demonstrate that Struct-Searcher is (1) plug-and-play and model-agnostic, yielding an average relative accuracy improvement of 17.2% on BrowseComp-VL across five different backbones. (2) top-performing, consistently outperforming state-of-the-artvision-language models(VLMs) anddeep research agents, with relative accuracy improvements of 3.7% on MM-BrowseComp, 1.5% on HLE-VL, and 0.7% on BrowseComp-VL over the second-best competing approach.

View arXiv pageView PDFAdd to collection

Models citing this paper0

No model linking this paper

Cite arxiv.org/abs/2606.07689 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2606.07689 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2606.07689 in a Space README.md to link it from this page.

Collections including this paper0

No Collection including this paper

Add this paper to acollectionto link it from this page.

Similar Articles

Agentic search models (5 minute read)

TLDR AI

Agentic search models are LLMs trained specifically for orchestrating search tasks, offering smaller, faster, and domain-specific alternatives to general models like GPT-5. They unbundle the traditional monolithic search stack by allowing an intelligent model to manage the entire retrieval process.

SearchEyes: Towards Frontier Multimodal Deep Search Intelligence via Search World Simulation

arXiv cs.AI

SearchEyes uses a typed knowledge graph as the backbone of a simulated search world to unify training data, search environment, and reward signals. It proposes Perception-Knowledge Chains (PKC) for multi-hop path sampling and Hop-Anchored Policy Optimization (HaPO) for step-level credit assignment, achieving state-of-the-art performance on multimodal knowledge-intensive benchmarks.