Sinkhorn divergence
PulseAugur coverage of Sinkhorn divergence — every cluster mentioning Sinkhorn divergence across labs, papers, and developer communities, ranked by signal.
-
New framework enhances neural network reconstruction of random fields
Researchers have introduced a new local Sinkhorn divergence framework designed for training stochastic neural networks (SNNs) to reconstruct multidimensional random fields. This framework utilizes debiased Sinkhorn dive…
-
New framework uses local Sinkhorn divergence for random field reconstruction
Researchers have introduced a new framework utilizing local Sinkhorn divergence for reconstructing conditional distributions in multidimensional random fields. This approach enables the training of stochastic neural net…
-
New GDRO Framework Enhances Generative Models for Robust Optimization
Researchers have introduced Generative Distributionally Robust Optimization (GDRO), a new framework for generative models in distributionally robust optimization. GDRO addresses limitations in existing methods by allowi…
-
New research explores convergence of graph Laplacians with symmetric divergence
Researchers have published a paper detailing the convergence properties of graph Laplacians when constructed with a symmetric divergence on Riemannian submanifolds. The work establishes a bound relating the symmetric di…
-
New method improves AI emulation of chaotic systems
Researchers have developed a new method for training machine learning emulators to better model chaotic dynamical systems. This approach utilizes adversarial optimal transport objectives to learn high-quality summary st…
-
New generative models leverage Wasserstein flows for faster, higher-quality outputs
Researchers are exploring new methods for generative modeling, focusing on Wasserstein gradient flows to improve efficiency and sample quality. One approach, W-Flow, achieves state-of-the-art one-step generation for ima…