WebClipper: Efficient Evolution of Web Agents with Graph-based Trajectory Pruning

TL;DR

WebClipper introduces a graph-based trajectory pruning method to enhance web agent efficiency, reducing redundant reasoning steps and tool-call rounds while maintaining accuracy.

Contribution

The paper presents a novel graph-based pruning framework that compresses web agent trajectories, improving search efficiency and introducing a new performance metric.

Findings

Reduces tool-call rounds by about 20%

Maintains high accuracy with pruned trajectories

Introduces the F-AE Score for performance evaluation

Abstract

Deep Research systems based on web agents have shown strong potential in solving complex information-seeking tasks, yet their search efficiency remains underexplored. We observe that many state-of-the-art open-source web agents rely on long tool-call trajectories with cyclic reasoning loops and exploration of unproductive branches. To address this, we propose WebClipper, a framework that compresses web agent trajectories via graph-based pruning. Concretely, we model the agent's search process as a state graph and cast trajectory optimization as a minimum-necessary Directed Acyclic Graph (DAG) mining problem, yielding pruned trajectories that preserve essential reasoning while eliminating redundant steps. Continued training on these refined trajectories enables the agent to evolve toward more efficient search patterns and reduces tool-call rounds by about 20% while improving accuracy. Furthermore, we introduce a new metric called F-AE Score to measure the model's overall performance in balancing accuracy and efficiency. Experiments demonstrate that WebClipper compresses tool-call rounds under excellent performance, providing practical insight into balancing effectiveness and efficiency in web agent design.