Beyond Distance: Quantifying Point Cloud Dynamics with Persistent Homology and Dynamic Optimal Transport

Collection of recent ML/AI research (Mar 19 2026) spanning: topological analysis of dynamic point clouds (TpOT extensions), theoretical and empirical analyses of transformer recall/storage under realistic embeddings, partial-label learning with adaptive k-NN, distributionally-robust feature selection under covariate shift, amortized deep adaptive experimental design for dynamical systems, doubly-robust tests for conditional distributional treatment effects, global optimization (HALO), quantum amplitude estimation for tail-risk pricing, formal guarantees for multi-evidence aggregation (Latent P