Associate Professor, University of California, Davis
3 papers at NeurIPS 2025
Changing the DenseAM kernel from the standard Gaussian kernel to the KDE-optimal Epanechnikov kernel results in 1) exponential capacity without the exponential and 2) meaningful, emergent memories.
We bound the restricted spectral gap of simulated tempering chain for mixture distributions and show that simulated tempering Metropolis–Hastings efficiently samples from Gaussian mixtures.
