Saved in:
Bibliographic Details
Main Authors: Wang, Xinjun, Wang, Shengyao, Zhou, Aimin, Hao, Hao
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2603.02626
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915830836494336
author Wang, Xinjun
Wang, Shengyao
Zhou, Aimin
Hao, Hao
author_facet Wang, Xinjun
Wang, Shengyao
Zhou, Aimin
Hao, Hao
contents Autonomous web navigation requires agents to perceive complex visual environments and maintain long-term context, yet current Large Language Model (LLM) based agents often struggle with spatial disorientation and navigation loops. In this paper, we propose generally applicable V-GEMS(Visual Grounding and Explicit Memory System), a robust multimodal agent architecture designed for precise and resilient web traversal. Our agent integrates visual grounding to resolve ambiguous interactive elements and introduces an explicit memory stack with state tracking. This dual mechanism allows the agent to maintain a structured map of its traversal path, enabling valid backtracking and preventing cyclical failures in deep navigation tasks. We also introduce an updatable dynamic benchmark to rigorously evaluate adaptability. Experiments show V-GEMS significantly dominates the WebWalker baseline, achieving a substantial 28.7% performance gain. Code is available at https://github.com/Vaultttttttttttt/V-GEMS.
format Preprint
id arxiv_https___arxiv_org_abs_2603_02626
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle See and Remember: A Multimodal Agent for Web Traversal
Wang, Xinjun
Wang, Shengyao
Zhou, Aimin
Hao, Hao
Artificial Intelligence
Autonomous web navigation requires agents to perceive complex visual environments and maintain long-term context, yet current Large Language Model (LLM) based agents often struggle with spatial disorientation and navigation loops. In this paper, we propose generally applicable V-GEMS(Visual Grounding and Explicit Memory System), a robust multimodal agent architecture designed for precise and resilient web traversal. Our agent integrates visual grounding to resolve ambiguous interactive elements and introduces an explicit memory stack with state tracking. This dual mechanism allows the agent to maintain a structured map of its traversal path, enabling valid backtracking and preventing cyclical failures in deep navigation tasks. We also introduce an updatable dynamic benchmark to rigorously evaluate adaptability. Experiments show V-GEMS significantly dominates the WebWalker baseline, achieving a substantial 28.7% performance gain. Code is available at https://github.com/Vaultttttttttttt/V-GEMS.
title See and Remember: A Multimodal Agent for Web Traversal
topic Artificial Intelligence
url https://arxiv.org/abs/2603.02626