AI RESEARCH
Stop Marginalizing My Dreams: Model Inversion via Laplace Kernel for Continual Learning
arXiv CS.LG
•
ArXi:2605.11804v1 Announce Type: new Data-free continual learning (DFCIL) relies on model inversion to synthesize pseudo-samples and mitigate catastrophic forgetting. However, existing inversion methods are fundamentally limited by a simplifying assumption: they model feature distributions using diagonal covariance, effectively ignoring correlations that define the geometry of learned representations. As a result, synthesized samples often lack fidelity, limiting knowledge retention. In this work, we show that modeling feature dependencies is a key ingredient for effective DFCIL. We