AI RESEARCH
Neural Collapse in Test-Time Adaptation
arXiv CS.CV
•
ArXi:2512.10421v2 Announce Type: replace Test-Time Adaptation (TTA) enhances model robustness to out-of-distribution (OOD) data by updating the model online during inference, yet existing methods lack theoretical insights into the fundamental causes of performance degradation under domain shifts. Recently, Neural Collapse (NC) has been proposed as an emergent geometric property of deep neural networks (DNNs), providing valuable insights for