Statistical learning theory and Occam's razor: Regularization

A collection of recent arXiv papers on statistical learning theory and machine learning, covering regularization and Occam’s razor, selective inference, manifold-aware diffusion models, density-ridge convergence, information-theoretic limits of deep learning, Wasserstein-manifold optimization, multimodal alignment, conformal prediction fairness, and multicalibration sample complexity. The material is academic and does not describe cybersecurity vulnerabilities, exploits, malicious activity, or affected software products.