AI & ML New Capability

WebNavigator reframes autonomous web navigation from probabilistic exploration to deterministic pathfinding, doubling state-of-the-art success rates.

March 24, 2026

Original Paper

WebNavigator: Global Web Navigation via Interaction Graph Retrieval

Xuanwang Zhang, Yuteng Han, Jinnan Qi, Mulong Xie, Zhen Wu, Xinyu Dai

arXiv · 2603.20366

The Takeaway

It solves 'Topological Blindness' by building interaction graphs offline and using a 'Retrieve-Reason-Teleport' workflow. This shift allows agents to navigate complex multi-site tasks with a 72.9% success rate, where previous agents struggled at under 35%.

From the abstract

Despite significant advances in autonomous web navigation, current methods remain far from human-level performance in complex web environments. We argue that this limitation stems from Topological Blindness, where agents are forced to explore via trial-and-error without access to the global topological structure of the environment. To overcome this limitation, we introduce WebNavigator, which reframes web navigation from probabilistic exploration into deterministic retrieval and pathfinding. Web