A locally-built, LLM-digested index of recent arXiv papers in quant finance, geometry/topology, and statistical ML — keyword search served straight from SQLite on this machine.
A new method uses higher-order Langevin dynamics with critical damping for better generative modeling.
TOLD++ improves convergence of diffusion models by critically damping the forward transition matrix.
Improved sampling in generative models using CLDs with a hyperparameter.
Improved generative models using critically-damped Langevin diffusion.
Introduces VSMD to improve generative diffusion processes without high costs.
New optimization method helps models generalize better after achieving near-perfect training performance.
Paper defends diffusion models from membership inference attacks using Langevin dynamics.
New model explains GAN training dynamics and mode collapse.
New non-separable covariance kernels for spatiotemporal data derived from harmonic oscillator physics.