AmbiBench: Benchmarking Mobile GUI Agents Beyond One-Shot Instructions in the Wild

TL;DR

AmbiBench introduces a new benchmark for mobile GUI agents that evaluates their ability to interpret and clarify ambiguous instructions through bidirectional interaction, advancing beyond traditional one-shot instruction following.

Contribution

This paper presents AmbiBench, the first benchmark incorporating instruction clarity taxonomy and a multi-agent evaluation framework to assess intent alignment in mobile GUI agents.

Findings

State-of-the-art agents struggle with ambiguous instructions.

Active clarification significantly improves task success.

MUSE correlates strongly with human judgment.

Abstract

Benchmarks are paramount for gauging progress in the domain of Mobile GUI Agents. In practical scenarios, users frequently fail to articulate precise directives containing full task details at the onset, and their expressions are typically ambiguous. Consequently, agents are required to converge on the user's true intent via active clarification and interaction during execution. However, existing benchmarks predominantly operate under the idealized assumption that user-issued instructions are complete and unequivocal. This paradigm focuses exclusively on assessing single-turn execution while overlooking the alignment capability of the agent. To address this limitation, we introduce AmbiBench, the first benchmark incorporating a taxonomy of instruction clarity to shift evaluation from unidirectional instruction following to bidirectional intent alignment. Grounded in Cognitive Gap theory, we propose a taxonomy of four clarity levels: Detailed, Standard, Incomplete, and Ambiguous. We construct a rigorous dataset of 240 ecologically valid tasks across 25 applications, subject to strict review protocols. Furthermore, targeting evaluation in dynamic environments, we develop MUSE (Mobile User Satisfaction Evaluator), an automated framework utilizing an MLLM-as-a-judge multi-agent architecture. MUSE performs fine-grained auditing across three dimensions: Outcome Effectiveness, Execution Quality, and Interaction Quality. Empirical results on AmbiBench reveal the performance boundaries of SoTA agents across different clarity levels, quantify the gains derived from active interaction, and validate the strong correlation between MUSE and human judgment. This work redefines evaluation standards, laying the foundation for next-generation agents capable of truly understanding user intent.