![[2602.14083] Plan-MCTS: Plan Exploration for Action Exploitation in Web Navigation](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2602.14083] Plan-MCTS: Plan Exploration for Action Exploitation in Web Navigation
Summary
The article presents Plan-MCTS, a novel framework for enhancing web navigation through improved exploration and state perception, addressing challenges faced by autonomous agents.
Why It Matters
As web navigation becomes increasingly complex, efficient exploration methods are crucial for autonomous agents. Plan-MCTS offers a significant advancement by transforming sparse action spaces into dense plan trees, improving task effectiveness and search efficiency, which can have broad implications for AI applications in various domains.
Key Takeaways
- Plan-MCTS reformulates web navigation by utilizing a semantic Plan Space.
- The framework enhances exploration efficiency by creating a Dense Plan Tree.
- It improves state awareness through an Abstracted Semantic History.
- Plan-MCTS incorporates a Dual-Gating Reward for validating actions.
- Extensive experiments show it outperforms existing methods in effectiveness and efficiency.
Computer Science > Artificial Intelligence arXiv:2602.14083 (cs) [Submitted on 15 Feb 2026] Title:Plan-MCTS: Plan Exploration for Action Exploitation in Web Navigation Authors:Weiming Zhang, Jihong Wang, Jiamu Zhou, Qingyao Li, Xinbei Ma, Congmin Zheng, Xingyu Lou, Weiwen Liu, Zhuosheng Zhang, Jun Wang, Yong Yu, Weinan Zhang View a PDF of the paper titled Plan-MCTS: Plan Exploration for Action Exploitation in Web Navigation, by Weiming Zhang and 11 other authors View PDF HTML (experimental) Abstract:Large Language Models (LLMs) have empowered autonomous agents to handle complex web navigation tasks. While recent studies integrate tree search to enhance long-horizon reasoning, applying these algorithms in web navigation faces two critical challenges: sparse valid paths that lead to inefficient exploration, and a noisy context that dilutes accurate state perception. To address this, we introduce Plan-MCTS, a framework that reformulates web navigation by shifting exploration to a semantic Plan Space. By decoupling strategic planning from execution grounding, it transforms sparse action space into a Dense Plan Tree for efficient exploration, and distills noisy contexts into an Abstracted Semantic History for precise state awareness. To ensure efficiency and robustness, Plan-MCTS incorporates a Dual-Gating Reward to strictly validate both physical executability and strategic alignment and Structural Refinement for on-policy repair of failed subplans. Extensive experiments on We...
