7 papers
Uncertainty Estimation and Generalization Bounds for Modern Deep Learning
Luis A. Ortega
This thesis investigates how Bayesian principles can deepen our understanding of modern deep learning systems. While neural networks achieve remarkable predictive performance, thei…
Flow-Transformed Implicit Processes for Function-Space Variational Inference
Luis A. Ortega, Andrés R. Masegosa, Thomas D. Nielsen
Implicit-process priors define distributions over functions through flexible generative mechanisms, making them attractive for Bayesian function-space modelling. However, performin…
Fixed-Mean Gaussian Processes for Post-hoc Bayesian Deep Learning
Luis A. Ortega, Simón RodrÃguez-Santana, Daniel Hernández-Lobato
Recently, there has been an increasing interest in performing post-hoc uncertainty estimation about the predictions of pre-trained deep neural networks (DNNs). Given a pre-trained…
Improving the Linearized Laplace Approximation via Quadratic Approximations
Pedro Jiménez, Luis A. Ortega, Pablo Morales-Ãlvarez +1
Deep neural networks (DNNs) often produce overconfident out-of-distribution predictions, motivating Bayesian uncertainty quantification. The Linearized Laplace Approximation (LLA)…
Scalable Linearized Laplace Approximation via Surrogate Neural Kernel
Luis A. Ortega, Simón RodrÃguez-Santana, Daniel Hernández-Lobato
We introduce a scalable method to approximate the kernel of the Linearized Laplace Approximation (LLA). For this, we use a surrogate deep neural network (DNN) that learns a compact…
PAC-Chernoff Bounds: Understanding Generalization in the Interpolation Regime
Andrés R. Masegosa, Luis A. Ortega
This paper introduces a distribution-dependent PAC-Chernoff bound that exhibits perfect tightness for interpolators, even within over-parameterized model classes. This bound, which…