AgentChangeBench: A Multi-Dimensional Evaluation Framework for Goal-Shift Robustness in Conversational AI

Explore AgentChangeBench, a rigorous framework for measuring how conversational AI adapts to shifting goals. This research highlights critical performance ga...

Level: advanced

By Manik Rana and 6 other authors

Category: research