Derives ODE limits of Adam-DA showing that first- and second-order momentum parameters reverse their convergence roles in zero-sum games compared to minimization, validated on GAN experiments.
Title resolution pending
6 Pith papers cite this work. Polarity classification is still indexing.
6
Pith papers citing it
representative citing papers
Domain transfer becomes identifiable from marginals plus one anchor under Jacobian sparsity, enabled by a randomized masked finite-difference regularizer.
Sobolev regularization on the witness function enables global convergence of MMD gradient flows for both sampling and generative modeling without isoperimetric assumptions.
Cardiac Mesh Flow generates 3D+t four-chamber cardiac meshes with anatomical correspondence and volume conditioning via one-step flow matching on multi-scale deformation fields.
Scion is a new stochastic LMO-based optimizer family that unifies existing methods, supports unconstrained problems, and delivers hyperparameter transferability plus speedups on nanoGPT training.
citing papers explorer
-
Understanding Dynamics of Adam in Zero-Sum Games: An ODE Approach
Derives ODE limits of Adam-DA showing that first- and second-order momentum parameters reverse their convergence roles in zero-sum games compared to minimization, validated on GAN experiments.
-
Domain Transfer Becomes Identifiable via a Single Alignment
Domain transfer becomes identifiable from marginals plus one anchor under Jacobian sparsity, enabled by a randomized masked finite-difference regularizer.
-
Sobolev Regularized MMD Gradient Flow
Sobolev regularization on the witness function enables global convergence of MMD gradient flows for both sampling and generative modeling without isoperimetric assumptions.
-
Cardiac Mesh Flow: One-Step Generation of 3D+t Cardiac Four-Chamber Meshes via Flow Matching
Cardiac Mesh Flow generates 3D+t four-chamber cardiac meshes with anatomical correspondence and volume conditioning via one-step flow matching on multi-scale deformation fields.
-
Training Deep Learning Models with Norm-Constrained LMOs
Scion is a new stochastic LMO-based optimizer family that unifies existing methods, supports unconstrained problems, and delivers hyperparameter transferability plus speedups on nanoGPT training.
- Reducing Diffusion Model Memorization with Higher Order Langevin Dynamics