HiconAgent: History Context-aware Policy Optimization for GUI Agents

TL;DR

HiconAgent introduces a novel history-aware policy optimization method for GUI agents, improving efficiency and accuracy in sequential navigation tasks by adaptively utilizing relevant historical context.

Contribution

The paper proposes HCPO with DCS and AHC components, enabling adaptive history sampling and compression, leading to better performance and efficiency in GUI navigation.

Findings

HiconAgent-3B outperforms larger models on GUI-Odyssey by +8.46% accuracy.

Achieves up to 2.47x speedup and 60% FLOPs reduction.

Demonstrates strong performance on multiple GUI benchmarks.

Abstract

Graphical User Interface (GUI) agents require effective use of historical context to perform sequential navigation tasks. While incorporating past actions and observations can improve decision making, naive use of full history leads to excessive computational overhead and distraction from irrelevant information. To address this, we introduce HiconAgent, a GUI agent trained with History Context-aware Policy Optimization (HCPO) for efficient and effective utilization of historical information. HCPO optimizes history usage in both sampling and policy updates through two complementary components: (1) Dynamic Context Sampling (DCS) presents the agent with variable length histories during sampling, enabling adaptive use of the most relevant context; (2) Anchor-guided History Compression (AHC) refines the policy update phase with a dual branch strategy where the compressed branch removes history observations while keeping history actions as information flow anchors. The compressed and uncompressed branches are coupled through a history-enhanced alignment loss to enforce consistent history usage while maintaining efficiency. Experiments on mainstream GUI navigation benchmarks demonstrate strong performance. Despite being smaller, HiconAgent-3B outperforms GUI-R1-7B by +8.46 percent grounding accuracy and +11.32 percent step success rate on GUI-Odyssey, while achieving comparable results on AndroidControl and AITW with up to 2.47x computational speedup and 60 percent FLOPs reduction.