Bayesian-Neural-Networks
Pytorch implementations of Bayes By Backprop, MC Dropout, SGLD, the Local Reparametrization Trick, KF-Laplace, SG-HMC and more
- 2k
- Jupyter Notebook
- MIT
- 3 y ago
Pytorch implementations of Bayes By Backprop, MC Dropout, SGLD, the Local Reparametrization Trick, KF-Laplace, SG-HMC and more
Open-source framework for uncertainty and deep learning models in PyTorch 🌱
[ICCV'19] Calibration Wizard: A Guidance System for Camera Calibration Based on Modelling Geometric and Corner Uncertainty
Official pytorch implementation of the paper "Bayesian Meta-Learning for the Few-Shot Setting via Deep Kernels" (NeurIPS 2020)
Pre-Training Buys Better Robustness and Uncertainty Estimates (ICML 2019)
Code for "Depth Uncertainty in Neural Networks" (https://arxiv.org/abs/2006.08437)
(ICML 2023) High Fidelity Image Counterfactuals with Probabilistic Causal Models
🔎Official code for our paper: "VL-Uncertainty: Detecting Hallucination in Large Vision-Language Model via Uncertainty Estimation".
Code for the paper "Getting a CLUE: A Method for Explaining Uncertainty Estimates"
An agile life cycle assessment (LCA) platform that simultaneously characterize the economic and environmental impacts of biofuels and bioproducts under uncertainty.
Source code of "Calibrating Large Language Models Using Their Generations Only", ACL2024
[ECCV 2026] Official implementation of the paper: "Schrödinger's Cat: Probabilistic Representation and Prediction of Potential Scene Kinematics"
Code for "Variational Depth Search in ResNets" (https://arxiv.org/abs/2002.02797)
Disentangling Sources of Uncertainty for Active Exploration (Reinforcement Learning)
Noise-Adaptive Driving Assistance System (NADAS) using Deep Reinforcement Learning, State-Estimation & State Representation
Building energy you can trace to its source: an interactive map of heating demand from open data, with ranges instead of single numbers. First city: Munich.
We show that a model owner can artificially introduce uncertainty into their model and provide a corresponding detection mechanism.
Invert LLM hallucination into productive features
Runs in the browser on Hugging Face
Text classification model, used with transformers.
See which words an LLM was unsure about, from logprobs
Runs in the browser on Hugging Face
Token-level uncertainty visualisation
Runs in the browser on Hugging Face
Researching policy-induced observability in recommender systems: when serving policies create self-confirming preference models, and when exploration is worth the cost.
Research-style evaluation of local LLMs focused on paired comparisons, uncertainty, benchmark stability, and trustworthy model-ranking conclusions.