Sharpness-aware minimization (SAM) was proposed to reduce sharpness of m...
Studying neural network loss landscapes provides insights into the natur...
We propose multirate training of neural networks: partitioning neural ne...
We employ constraints to control the parameter space of deep neural netw...
We propose a method for efficiently incorporating constraints into a
sto...
Stochastic Gradient Langevin Dynamics, the "unadjusted Langevin algorith...