WebSynthesis: World-Model-Guided MCTS for Efficient WebUI-Trajectory Synthesis

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Gao, Yifei, Ye, Junhong, Wang, Jiaqi, Sang, Jitao
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911041733001216
author Gao, Yifei
Ye, Junhong
Wang, Jiaqi
Sang, Jitao
author_facet Gao, Yifei
Ye, Junhong
Wang, Jiaqi
Sang, Jitao
contents Recent advancements in large language models (LLMs) have significantly improved the capabilities of web agents. However, effectively navigating complex and dynamic web environments still requires more advanced trajectory-level planning and execution. Prior studies have addressed self-improving agents by collecting extensive GUI trajectories from real-environment interactions. Despite their effectiveness, these approaches encounter two critical challenges: (1) Uncontrollable environment states, where real or sandboxed web environments often yield unstable and non-deterministic feedback, complicating the reproduction and debugging of agent behaviors; and (2) High API costs, as generating even a single interaction trajectory can involve hundreds of queries, leading to considerable API usage and computational expenses. To address these limitations and enable scalable self-improvement for agents, we propose WebSynthesis, a novel framework for trajectory synthesis and training. WebSynthesis leverages a learned world model to simulate virtual web environments, allowing a policy agent to perform efficient and reversible tree-based planning. This approach supports the large-scale generation of diverse and high-quality trajectories, which are subsequently utilized to refine the agent's policy. Experimental results demonstrate that an agent trained using WebSynthesis on a small-scale synthetic dataset achieves performance comparable to or even surpassing that of models trained on large-scale real-world data.
format Preprint
id arxiv_https___arxiv_org_abs_2507_04370
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle WebSynthesis: World-Model-Guided MCTS for Efficient WebUI-Trajectory Synthesis
Gao, Yifei
Ye, Junhong
Wang, Jiaqi
Sang, Jitao
Artificial Intelligence
Recent advancements in large language models (LLMs) have significantly improved the capabilities of web agents. However, effectively navigating complex and dynamic web environments still requires more advanced trajectory-level planning and execution. Prior studies have addressed self-improving agents by collecting extensive GUI trajectories from real-environment interactions. Despite their effectiveness, these approaches encounter two critical challenges: (1) Uncontrollable environment states, where real or sandboxed web environments often yield unstable and non-deterministic feedback, complicating the reproduction and debugging of agent behaviors; and (2) High API costs, as generating even a single interaction trajectory can involve hundreds of queries, leading to considerable API usage and computational expenses. To address these limitations and enable scalable self-improvement for agents, we propose WebSynthesis, a novel framework for trajectory synthesis and training. WebSynthesis leverages a learned world model to simulate virtual web environments, allowing a policy agent to perform efficient and reversible tree-based planning. This approach supports the large-scale generation of diverse and high-quality trajectories, which are subsequently utilized to refine the agent's policy. Experimental results demonstrate that an agent trained using WebSynthesis on a small-scale synthetic dataset achieves performance comparable to or even surpassing that of models trained on large-scale real-world data.
title WebSynthesis: World-Model-Guided MCTS for Efficient WebUI-Trajectory Synthesis
topic Artificial Intelligence
url https://arxiv.org/abs/2507.04370