Beyond Generative Priors: Minority Sampling with JEPA-Guided Diffusion 事件

Beyond Generative Priors: Minority Sampling with JEPA-Guided Diffusion arXiv:2605.24631v1 Announce Type: cross Abstract: Minority sampling aims to generate low-density instances on a data manifold and is of central importance in applications such as medical diagnosis, anomaly detection, and creative AI. Existing approaches, however, define minority samples relative to generative priors learned from training data, confining rarity to model-specific notions that may poorly reflect real-world sema