InfiGUI-G1: Advancing GUI Grounding with Adaptive Exploration Policy Optimization

TL;DR

This paper introduces AEPO, a new policy optimization framework that enhances GUI grounding by improving semantic alignment through broader exploration, leading to state-of-the-art results on multiple benchmarks.

Contribution

The paper proposes AEPO, an innovative exploration strategy with a theoretically grounded reward, significantly advancing semantic alignment in GUI grounding tasks.

Findings

Achieved up to 9.0% improvement over baseline

Established new state-of-the-art on multiple benchmarks

Demonstrated effectiveness of broader exploration in semantic tasks

Abstract

The emergence of Multimodal Large Language Models (MLLMs) has propelled the development of autonomous agents that operate on Graphical User Interfaces (GUIs) using pure visual input. A fundamental challenge is robustly grounding natural language instructions. This requires a precise spatial alignment, which accurately locates the coordinates of each element, and, more critically, a correct semantic alignment, which matches the instructions to the functionally appropriate UI element. Although Reinforcement Learning with Verifiable Rewards (RLVR) has proven to be effective at improving spatial alignment for these MLLMs, we find that inefficient exploration bottlenecks semantic alignment, which prevent models from learning difficult semantic associations. To address this exploration problem, we present Adaptive Exploration Policy Optimization (AEPO), a new policy optimization framework. AEPO employs a multi-answer generation strategy to enforce broader exploration, which is then guided by a theoretically grounded Adaptive Exploration Reward (AER) function derived from first principles of efficiency eta=U/C. Our AEPO-trained models, InfiGUI-G1-3B and InfiGUI-G1-7B, establish new state-of-the-art results across multiple challenging GUI grounding benchmarks, achieving significant relative improvements of up to 9.0% against the naive RLVR baseline on benchmarks designed to test generalization and semantic understanding. Resources are available at https://github.com/InfiXAI/InfiGUI-G1.