AndroTMem: From Interaction Trajectories to Anchored Memory in Long-Horizon GUI Agents

2026-03-18

AI Summary

This work addresses the degraded task completion performance of existing long-horizon GUI agents, which often stems from inefficient memory mechanisms leading to redundant replays or loss of critical dependency information. The study presents the first systematic diagnosis of this memory failure and introduces Anchored State Memory (ASM), a novel approach that compresses interaction trajectories into structured intermediate state anchors through causal dependency analysis, thereby enabling subgoal retrieval and attribution-aware decision-making. To support this investigation, the authors develop the AndroTMem diagnostic framework and a new benchmark, AndroTMem-Bench, comprising 1,069 tasks with strong causal dependencies. Experiments across 12 GUI agents demonstrate that ASM consistently outperforms full-sequence replay and summarization baselines, improving task completion rates by 5%–30.16% and average memory scores by 4.93%–24.66%.

Technology Category

• Multiagent Systems: Adversarial Agents
• Cognitive Modeling & Cognitive Systems: Agent Architectures
• Humans and AI: Human-Aware Planning and Behavior Prediction

Application Category

• Search and Retrieval-Augmented AI: Agentic search
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

Long-horizon GUI agents are a key step toward real-world deployment, yet effective interaction memory under prevailing paradigms remains under-explored. Replaying full interaction sequences is redundant and amplifies noise, while summaries often erase dependency-critical information and traceability. We present AndroTMem, a diagnostic framework for anchored memory in long-horizon Android GUI agents. Its core benchmark, AndroTMem-Bench, comprises 1,069 tasks with 34,473 interaction steps (avg. 32.1 per task, max. 65). We evaluate agents with TCR (Task Complete Rate), focusing on tasks whose completion requires carrying forward critical intermediate state; AndroTMem-Bench is designed to enforce strong step-to-step causal dependencies, making sparse yet essential intermediate states decisive for downstream actions and centering interaction memory in evaluation. Across open- and closed-source GUI agents, we observe a consistent pattern: as interaction sequences grow longer, performance drops are driven mainly by within-task memory failures, not isolated perception errors or local action mistakes. Guided by this diagnosis, we propose Anchored State Memory (ASM), which represents interaction sequences as a compact set of causally linked intermediate-state anchors to enable subgoal-targeted retrieval and attribution-aware decision making. Across multiple settings and 12 evaluated GUI agents, ASM consistently outperforms full-sequence replay and summary-based baselines, improving TCR by 5%-30.16% and AMS by 4.93%-24.66%, indicating that anchored, structured memory effectively mitigates the interaction-memory bottleneck in long-horizon GUI tasks. The code, benchmark, and related resources are publicly available at [https://github.com/CVC2233/AndroTMem](https://github.com/CVC2233/AndroTMem).

Problem

• long-horizon GUI agents
• interaction memory
• intermediate state
• causal dependencies
• memory bottleneck

Innovation

• anchored memory
• long-horizon GUI agents
• interaction trajectories
• causal dependencies
• Anchored State Memory (ASM)