Last Layer Hamiltonian Monte Carlo

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Vellenga, Koen, Steinhauer, H. Joe, Falkman, Göran, Andersson, Jonas, Sjögren, Anders
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866911052630851584
author Vellenga, Koen
Steinhauer, H. Joe
Falkman, Göran
Andersson, Jonas
Sjögren, Anders
author_facet Vellenga, Koen
Steinhauer, H. Joe
Falkman, Göran
Andersson, Jonas
Sjögren, Anders
contents We explore the use of Hamiltonian Monte Carlo (HMC) sampling as a probabilistic last layer approach for deep neural networks (DNNs). While HMC is widely regarded as a gold standard for uncertainty estimation, the computational demands limit its application to large-scale datasets and large DNN architectures. Although the predictions from the sampled DNN parameters can be parallelized, the computational cost still scales linearly with the number of samples (similar to an ensemble). Last layer HMC (LL--HMC) reduces the required computations by restricting the HMC sampling to the final layer of a DNN, making it applicable to more data-intensive scenarios with limited computational resources. In this paper, we compare LL-HMC against five last layer probabilistic deep learning (LL-PDL) methods across three real-world video datasets for driver action and intention. We evaluate the in-distribution classification performance, calibration, and out-of-distribution (OOD) detection. Due to the stochastic nature of the probabilistic evaluations, we performed five grid searches for different random seeds to avoid being reliant on a single initialization for the hyperparameter configurations. The results show that LL--HMC achieves competitive in-distribution classification and OOD detection performance. Additional sampled last layer parameters do not improve the classification performance, but can improve the OOD detection. Multiple chains or starting positions did not yield consistent improvements.
format Preprint
id arxiv_https___arxiv_org_abs_2507_08905
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Last Layer Hamiltonian Monte Carlo
Vellenga, Koen
Steinhauer, H. Joe
Falkman, Göran
Andersson, Jonas
Sjögren, Anders
Machine Learning
Artificial Intelligence
G.3
We explore the use of Hamiltonian Monte Carlo (HMC) sampling as a probabilistic last layer approach for deep neural networks (DNNs). While HMC is widely regarded as a gold standard for uncertainty estimation, the computational demands limit its application to large-scale datasets and large DNN architectures. Although the predictions from the sampled DNN parameters can be parallelized, the computational cost still scales linearly with the number of samples (similar to an ensemble). Last layer HMC (LL--HMC) reduces the required computations by restricting the HMC sampling to the final layer of a DNN, making it applicable to more data-intensive scenarios with limited computational resources. In this paper, we compare LL-HMC against five last layer probabilistic deep learning (LL-PDL) methods across three real-world video datasets for driver action and intention. We evaluate the in-distribution classification performance, calibration, and out-of-distribution (OOD) detection. Due to the stochastic nature of the probabilistic evaluations, we performed five grid searches for different random seeds to avoid being reliant on a single initialization for the hyperparameter configurations. The results show that LL--HMC achieves competitive in-distribution classification and OOD detection performance. Additional sampled last layer parameters do not improve the classification performance, but can improve the OOD detection. Multiple chains or starting positions did not yield consistent improvements.
title Last Layer Hamiltonian Monte Carlo
topic Machine Learning
Artificial Intelligence
G.3
url https://arxiv.org/abs/2507.08905