WebNavigator: Global Web Navigation via Interaction Graph Retrieval

TL;DR

WebNavigator introduces a novel approach to autonomous web navigation by constructing interaction graphs offline and using a retrieve-reason-teleport workflow, significantly improving success rates over previous methods.

Contribution

It presents WebNavigator, a new method that overcomes topological blindness by leveraging offline graph construction and deterministic pathfinding for better web navigation.

Findings

Achieves 72.9% success rate on WebArena tasks.

More than doubles the performance of existing agents.

Highlights topological blindness as a key bottleneck.

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. WebNavigator constructs Interaction Graphs via zero-token cost heuristic exploration offline and implements a Retrieve-Reason-Teleport workflow for global navigation online. WebNavigator achieves state-of-the-art performance on WebArena and OnlineMind2Web. On WebArena multi-site tasks, WebNavigator achieves a 72.9\% success rate, more than doubling the performance of enterprise-level agents. This work reveals that Topological Blindness, rather than model reasoning capabilities alone, is an underestimated bottleneck in autonomous web navigation.