Research map / Probabilistic and scientific modeling / Uncertainty and Gaussian processes
Feature Preserving Shrinkage on Bayesian Neural Networks Via the R2D2 Prior
Tsai Hor Chan, Ding Zhang, Guosheng Yin, Lequan Yu
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 variational Gibbs inference instead of SGMCMC with the R2D2 prior (ablation) | R2D2-Net · CIFAR-10, AlexNet | accuracy: 55.37 -> 65.10 | Improved |
| use variational Gibbs inference instead of SGMCMC with the R2D2 prior (ablation) | R2D2-Net · CIFAR-100, AlexNet | accuracy: 20.18 -> 36.12 | Improved |
| use variational Gibbs inference instead of SGLD with the R2D2 prior (ablation) | R2D2-Net · Simulation S1 (L=3) | MSE: 20.61 -> 8.81 | Improved |
| use R2D2 prior with SVGI instead of Gaussian prior with SVI | BNN · CIFAR-10, AlexNet | accuracy: 60.28 -> 65.10 | Improved |
| use R2D2 prior instead of horseshoe prior | BNN · CIFAR-100, AlexNet | accuracy: 33.27 -> 36.12 | Improved |
| use R2D2 prior instead of horseshoe prior | BNN · CIFAR-10, AlexNet | accuracy: 65.01 -> 65.10 | No clear effect |
| use R2D2 prior instead of horseshoe prior in OOD detection | BNN · CIFAR-10 vs Fashion MNIST | AUROC: 80.76 -> 92.85 | Improved |
| use the smaller LeNet architecture instead of AlexNet for R2D2-Net (ablation) | R2D2-Net · CIFAR-10 | accuracy vs frequentist: 62.14 vs 62.24 frequentist | No clear effect |
| increase MLP depth from L=0 to L=3 for R2D2-Net (ablation) | R2D2-Net · Simulation S1 | MSE: 414.36 -> 8.81 | Improved |
| vary hyperparameters a_pi, b and rho_0 of the R2D2 prior (ablation) | R2D2-Net · Simulation S1-S3 (L=3) | test MSE | No clear effect |
Closest papers by meaning
- Encoding the Latent Posterior of Bayesian Neural Networks for Uncertainty Quantification (TPAMI 2023 · similarity 0.88)
- Deterministic Variational Inference for Robust Bayesian Neural Networks (ICLR 2019 · similarity 0.88)
- Training Bayesian Neural Networks with Sparse Subspace Variational Inference (ICLR 2024 · similarity 0.88)
- A Study of Bayesian Neural Network Surrogates for Bayesian Optimization (ICLR 2024 · similarity 0.86)
- Variational Bayesian Last Layers (ICLR 2024 · similarity 0.86)
- Parameter Expanded Stochastic Gradient Markov Chain Monte Carlo (ICLR 2025 · similarity 0.86)
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.
