Saved in:
| Main Authors: | , , , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2604.00892 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866911559952891904 |
|---|---|
| author | Zou, Henry Peng Miao, Chunyu Huang, Wei-Chieh Chen, Yankai Zhou, Yue Zhang, Hanrong Wu, Yaozu Fang, Liancheng Gu, Zhengyao Zhang, Zhen Zheng, Kening Wang, Fangxin Nian, Yi Li, Shanghao Fan, Wenzhe He, Langzhou Zhang, Weizhi Liu, Xue Yu, Philip S. |
| author_facet | Zou, Henry Peng Miao, Chunyu Huang, Wei-Chieh Chen, Yankai Zhou, Yue Zhang, Hanrong Wu, Yaozu Fang, Liancheng Gu, Zhengyao Zhang, Zhen Zheng, Kening Wang, Fangxin Nian, Yi Li, Shanghao Fan, Wenzhe He, Langzhou Zhang, Weizhi Liu, Xue Yu, Philip S. |
| contents | 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. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_00892 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | When Users Change Their Mind: Evaluating Interruptible Agents in Long-Horizon Web Navigation Zou, Henry Peng Miao, Chunyu Huang, Wei-Chieh Chen, Yankai Zhou, Yue Zhang, Hanrong Wu, Yaozu Fang, Liancheng Gu, Zhengyao Zhang, Zhen Zheng, Kening Wang, Fangxin Nian, Yi Li, Shanghao Fan, Wenzhe He, Langzhou Zhang, Weizhi Liu, Xue Yu, Philip S. Computation and Language 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. |
| title | When Users Change Their Mind: Evaluating Interruptible Agents in Long-Horizon Web Navigation |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2604.00892 |