# JavierAntoran/Bayesian-Neural-Networks

Pytorch implementations of Bayes By Backprop, MC Dropout, SGLD, the Local Reparametrization Trick, KF-Laplace, SG-HMC and more

Repository: https://github.com/JavierAntoran/Bayesian-Neural-Networks
Canonical: https://ross.abutalabs.com/products/bayesian-neural-networks
Language: Jupyter Notebook
License: MIT
License Family: permissive
Topics: deep-learning, bayesian-inference, pytorch, reproducible-research, bayesian-neural-networks, bayes-by-backprop, mc-dropout, mcmc, variational-inference, kronecker-factored-approximation, langevin-dynamics, approximate-inference, local-reparametrization-trick, uncertainty, uncertainty-neural-networks, out-of-distribution-detection, classification, regression, sgld, hmc
Last push: 2023-10-20T01:10:33+00:00

## Health v2 (maintenance only)
Score: 32/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 2732, "days_push": 1049, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1969, forks 306 (observed 2026-08-28T04:06:00.727631+00:00)

## What it is
A collection of PyTorch implementations of Bayesian neural network approximate inference methods, including Bayes by Backprop, MC Dropout, SGLD, KF-Laplace, and SG-HMC. It ships as Jupyter notebooks and scripts with regression and MNIST classification experiments demonstrating uncertainty estimation.

## Use cases
- estimate neural network prediction uncertainty
- compare Bayesian deep learning inference methods
- run MC dropout experiments in pytorch
- implement Bayes by Backprop for regression
- detect out-of-distribution inputs with uncertainty
- reproduce Bayesian neural network research results
- train SGLD or SG-HMC models on MNIST

## When to choose
- you want reference implementations of multiple Bayesian inference methods in PyTorch
- you are studying or benchmarking uncertainty estimation techniques
- you need runnable notebooks for regression and classification uncertainty experiments

## When to avoid
- you need a production-ready, pip-installable library with a stable API
- you require Python 3 support and modern PyTorch versions out of the box
- you need scalable Bayesian inference for large models or datasets

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, deep-learning
- domain: deep-learning, machine-learning, artificial-intelligence
- platform: python, cross-platform
- tags: bayesian-neural-networks, pytorch, uncertainty-quantification, variational-inference, mc-dropout, sgld, bayes-by-backprop, laplace-approximation, research-code, jupyter-notebooks, gpu

## Member repositories
- JavierAntoran/Bayesian-Neural-Networks (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:06:00.727631+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T03:05:08.757542+00:00, confidence not recorded.
  - readme: https://github.com/JavierAntoran/Bayesian-Neural-Networks (fetched 2026-08-28T04:06:00.727631+00:00, sha e3fd40eb44e3)
- Data as of 2026-08-30T08:39:29.467469+00:00.