Tag
This paper develops a geometric dynamical framework to model how predictive AI assistance alters exploratory cognition by stabilizing trajectories before self-generated exploration, leading to reduced exploratory responsiveness, hysteresis, and delayed recovery upon assistance withdrawal.