AIMar 7

Enhancing Web Agents with a Hierarchical Memory Tree

arXiv:2603.07024v13 citations
Predicted impact top 33% in AI · last 90 daysOriginality Highly original
AI Analysis

This work addresses the problem of poor generalization of web agents across unseen websites, which is a significant challenge for developing robust automation tools.

The authors propose Hierarchical Memory Tree (HMT), a structured memory framework for web agents that decouples logical planning from action execution. HMT constructs a three-level hierarchy from raw trajectories, enabling a stage-aware inference mechanism that significantly outperforms flat-memory methods in cross-website and cross-domain scenarios on Mind2Web and WebArena.

Large language model-based web agents have shown strong potential in automating web interactions through advanced reasoning and instruction following. While retrieval-based memory derived from historical trajectories enables these agents to handle complex, long-horizon tasks, current methods struggle to generalize across unseen websites. We identify that this challenge arises from the flat memory structures that entangle high-level task logic with site-specific action details. This entanglement induces a workflow mismatch in new environments, where retrieved contents are conflated with current web, leading to logically inconsistent execution. To address this, we propose Hierarchical Memory Tree (HMT), a structured framework designed to explicitly decouple logical planning from action execution. HMT constructs a three-level hierarchy from raw trajectories via an automated abstraction pipeline: the Intent level maps diverse user instructions to standardized task goals; the Stage level defines reusable semantic subgoals characterized by observable pre-conditions and post-conditions; and the Action level stores action patterns paired with transferable semantic element descriptions. Leveraging this structure, we develop a stage-aware inference mechanism comprising a Planner and an Actor. By explicitly validating pre-conditions, the Planner aligns the current state with the correct logical subgoal to prevent workflow mismatch, while the Actor grounds actions by matching the stored semantic descriptions to the target page. Experimental results on Mind2Web and WebArena show that HMT significantly outperforms flat-memory methods, particularly in cross-website and cross-domain scenarios, highlighting the necessity of structured memory for robust generalization of web agents.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes