Publications

Publications


EvalAgent: Interactive Comparative Evaluation of Computer-Using GUI Agents

Published in IUI 2026 Companion, 2026

An interactive system for comparing computer-using GUI agents across models, personas, criteria, and full execution traces.

Recommended citation: Yang, Y., Gebreegziabher, S. A., Yoo, H., Chiang, C., Chen, C., Szymanski, A., Do, H. J., Ashktorab, Z., Geyer, W., Gómez-Zará, D., Li, T. J.-J. (2026). EvalAgent: Interactive Comparative Evaluation of Computer-Using GUI Agents. Companion Proceedings of IUI 2026, pp. 171-175.

The Behavioral Fabric of LLM-Powered GUI Agents: Human Values and Interaction Outcomes

Published in IUI, 2026

We investigate how users preferences and values impact LLM-powered web GUI agents reasoning and behavior across shopping, travel, dining, and housing tasks.

Recommended citation: Gebreegziabher, S., Yang, Y., Chiang, C., Yoo, H., Chen, C., Do, H. J., Ashktorab, Z., Geyer, W., Gomez-Zara, D., Li, T. J.-J. (2026). The Behavioral Fabric of LLM-Powered GUI Agents: Human Values and Interaction Outcomes. IUI 2026, pp. 909-927.