Research map / Probabilistic and scientific modeling / Uncertainty and Gaussian processes
Bayesian Deep Learning and a Probabilistic Perspective of Generalization
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| preprocess with randomly initialized untrained CNN | Gaussian kernel classifier · CIFAR-10 | accuracy: 54% -> 71% | Helped |
| multi-basin marginalization (MultiSWAG) instead of SGD | ResNet-18 varying width · CIFAR-100 (incl. 20% label noise) | test error / NLL | Helped |
| single-basin SWAG instead of SGD | ResNet-18 · CIFAR-100 20% label noise | test error | Helped |
| MultiSWAG instead of deep ensembles | PreResNet-20 · CIFAR-10 corrupted (Gaussian blur) | NLL | Helped |
| increase number of independent SWAG models in MultiSWAG (ablation) | MultiSWAG ResNet-18 · CIFAR-100 corrupted | test error | Helped |
| deep ensembles instead of single-basin VI | BNN regression · synthetic regression vs HMC | Wasserstein distance to true predictive | Helped |
Closest papers by meaning
- Learning Expressive Priors for Generalization and Uncertainty Estimation in Neural Networks (ICML 2023 · similarity 0.89)
- A Bayesian Perspective on Training Speed and Model Selection (NeurIPS 2020 · similarity 0.89)
- A Bayesian Perspective on Generalization and Stochastic Gradient Descent (ICLR 2018 · similarity 0.89)
- Bayesian Model Selection, the Marginal Likelihood, and Generalization (ICML 2022 · similarity 0.89)
- Bayesian Uncertainty Estimation for Batch Normalized Deep Networks (ICML 2018 · similarity 0.89)
- Bayesian Low-Rank Learning (Bella): A Practical Approach to Bayesian Neural Networks (AAAI 2025 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.