When Users Change Their Mind: Evaluating Interruptible Agents in Long-Horizon Web Navigation

TL;DR

This paper introduces a new benchmark and systematic evaluation of large language model agents' ability to handle user interruptions during complex web navigation tasks, highlighting current challenges.

Contribution

It formalizes realistic interruption types, creates the InterruptBench benchmark, and evaluates LLMs' effectiveness and efficiency in managing mid-task changes.

Findings

Handling interruptions remains challenging for large LLMs.

Evaluation across six models shows varied effectiveness in adaptation.

Efficiency in recovering from interruptions is limited.

Abstract

As LLM agents transition from short, static problem solving to executing complex, long-horizon tasks in dynamic environments, the ability to handle user interruptions, such as adding requirement or revising goals, during mid-task execution is becoming a core requirement for realistic deployment. However, existing benchmarks largely assume uninterrupted agent behavior or study interruptions only in short, unconstrained language tasks. In this paper, we present the first systematic study of interruptible agents in long-horizon, environmentally grounded web navigation tasks, where actions induce persistent state changes. We formalize three realistic interruption types, including addition, revision, and retraction, and introduce InterruptBench, a benchmark derived from WebArena-Lite that synthesizes high-quality interruption scenarios under strict semantic constraints. Using a unified interruption simulation framework, we evaluate six strong LLM backbones across single- and multi-turn interruption settings, analyzing both their effectiveness in adapting to updated intents and their efficiency in recovering from mid-task changes. Our results show that handling user interruptions effectively and efficiently during long-horizon agentic tasks remains challenging for powerful large-scale LLMs. Code and dataset are available at https://github.com/HenryPengZou/InterruptBench.