Research map / Probabilistic and scientific modeling / Uncertainty and Gaussian processes
Bayesian Deep Learning via Subnetwork Inference
Erik Daxberger, Eric Nalisnick, James U Allingham, Javier Antoran, Jose Miguel Hernandez-Lobato
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 |
|---|---|---|---|
| Use Wasserstein-selected rather than uniformly random subnetwork weights (ablation) | Subnetwork Linearized Laplace · MNIST vs FashionMNIST | AUC-ROC: 0.75 ± 0.02 -> 0.92 ± 0.05 | Improved |
| Use Wasserstein-selected rather than uniformly random subnetwork weights (ablation) | Subnetwork Linearized Laplace · CIFAR10 vs SVHN | AUC-ROC: 0.86 ± 0.02 -> 0.85 ± 0.03 | No clear effect |
| Use Wasserstein-selected rather than uniformly random subnetwork weights (ablation) | Subnetwork Linearized Laplace · MNIST, 45° rotation | log-likelihood: -1.60 \scriptstyle \pm 0.05 -> -1.09 \scriptstyle \pm 0.03 | Improved |
| Use Wasserstein-selected rather than uniformly random subnetwork weights (ablation) | Subnetwork Linearized Laplace · MNIST, 60° rotation | log-likelihood: -3.85 \scriptstyle \pm 0.18 -> -2.10 \scriptstyle \pm 0.03 | Improved |
| Use subnetwork inference instead of VOGN | ResNet-18 image classification · MNIST, 45° rotation | log-likelihood: -1.15 \scriptstyle \pm nan -> -1.09 \scriptstyle \pm 0.03 | Improved |
| Use subnetwork inference instead of VOGN | ResNet-18 image classification · MNIST, 60° rotation | log-likelihood: -1.92 \scriptstyle \pm nan -> -2.10 \scriptstyle \pm 0.03 | Hurt |
Closest papers by meaning
- Global inducing point variational posteriors for Bayesian neural networks and deep Gaussian processes (ICML 2021 · similarity 0.89)
- On the Expressiveness of Approximate Inference in Bayesian Neural Networks (NeurIPS 2020 · similarity 0.89)
- Liberty or Depth: Deep Bayesian Neural Nets Do Not Need Complex Weight Posterior Approximations (NeurIPS 2020 · similarity 0.89)
- A Symmetry-Aware Exploration of Bayesian Neural Network Posteriors (ICLR 2024 · similarity 0.88)
- A Study of Bayesian Neural Network Surrogates for Bayesian Optimization (ICLR 2024 · similarity 0.88)
- VIKING: Deep variational inference with stochastic projections (NeurIPS 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.
