Approximate Bayesian Inference via Bitstring Representations

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Sladek, Aleksanteri, Trapp, Martin, Solin, Arno
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866918127185428480
author Sladek, Aleksanteri
Trapp, Martin
Solin, Arno
author_facet Sladek, Aleksanteri
Trapp, Martin
Solin, Arno
contents The machine learning community has recently put effort into quantized or low-precision arithmetics to scale large models. This paper proposes performing probabilistic inference in the quantized, discrete parameter space created by these representations, effectively enabling us to learn a continuous distribution using discrete parameters. We consider both 2D densities and quantized neural networks, where we introduce a tractable learning approach using probabilistic circuits. This method offers a scalable solution to manage complex distributions and provides clear insights into model behavior. We validate our approach with various models, demonstrating inference efficiency without sacrificing accuracy. This work advances scalable, interpretable machine learning by utilizing discrete approximations for probabilistic computations.
format Preprint
id arxiv_https___arxiv_org_abs_2508_13598
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Approximate Bayesian Inference via Bitstring Representations
Sladek, Aleksanteri
Trapp, Martin
Solin, Arno
Machine Learning
The machine learning community has recently put effort into quantized or low-precision arithmetics to scale large models. This paper proposes performing probabilistic inference in the quantized, discrete parameter space created by these representations, effectively enabling us to learn a continuous distribution using discrete parameters. We consider both 2D densities and quantized neural networks, where we introduce a tractable learning approach using probabilistic circuits. This method offers a scalable solution to manage complex distributions and provides clear insights into model behavior. We validate our approach with various models, demonstrating inference efficiency without sacrificing accuracy. This work advances scalable, interpretable machine learning by utilizing discrete approximations for probabilistic computations.
title Approximate Bayesian Inference via Bitstring Representations
topic Machine Learning
url https://arxiv.org/abs/2508.13598