← Search

Michał Jamroż

2 accepted papers

2023

Neural Representations Reveal Distinct Modes of Class Fitting in Residual Convolutional Networks

AAAI 2023technical

We leverage probabilistic models of neural representations to investigate how residual networks fit classes. To this end, we estimate class-conditional density models for representations learned by deep ResNets. We then use these models to characterize distributions of representations across learned…

2020

A Bayesian Nonparametrics View into Deep Representations

NeurIPS 2020poster

We investigate neural network representations from a probabilistic perspective. Specifically, we leverage Bayesian nonparametrics to construct models of neural activations in Convolutional Neural Networks (CNNs) and latent representations in Variational Autoencoders (VAEs). This allows us to formula…