Liam Hodgkinson

Lecturer@University of Melbourne·澳大利亚·OpenReview

研究方向

deep learning · stochastic gradient descent · Monte Carlo methods · bayesian inference · probabilistic machine learning · interacting particle systems · metapopulation models · central limit theorems

6.0

合作者 (11)

Liam Hodgkinson

Uncertainty-Aware Diagnostics for Physics-Informed Machine Learning

Spectral Estimation with Free Decompression

Determinant Estimation under Memory Constraints and Neural Scaling Laws

Uncertainty Quantification with the Empirical Neural Tangent Kernel

Models of Heavy-Tailed Mechanistic Universality

Temperature Optimization for Bayesian Deep Learning