Research map / Probabilistic and scientific modeling / Uncertainty and Gaussian processes
Variational Bayesian Dropout With a Hierarchical Prior
Check 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 |
|---|---|---|---|
| replace VD's improper log-uniform weight prior with a Gaussian prior whose variance has a uniform hyper-prior | Variational dropout · MNIST | test error (%) | Helped |
| replace VD's improper log-uniform weight prior with a Gaussian prior whose variance has a uniform hyper-prior | Variational dropout · CIFAR-10 | test error (%) | Helped |
| replace VD's improper log-uniform weight prior with a Gaussian prior whose variance has a uniform hyper-prior | Variational dropout · SVHN | test error (%) | Helped |
| use VBD rather than VD for weight compression | LeNet-300-100 · MNIST | test error (%): 1.94 -> 1.67 | Helped |
| use VBD high-sparsity variant rather than VD for weight compression | LeNet-300-100 · MNIST | compression ratio: 68 -> 81 | Helped |
| use VBD low-test-error variant rather than VD for weight compression | LeNet-5-Caffe · MNIST | test error (%): 0.75 -> 0.74 | Helped |
| use VBD high-sparsity variant rather than VD for weight compression | LeNet-5-Caffe · MNIST | compression ratio: 280 -> 290 | Helped |
| use VBD rather than VD for structured compression | LeNet-500-300 · MNIST | test error (%): 1.57 -> 1.35 | Helped |
| use VBD rather than VD for structured compression | LeNet-5-Caffe · MNIST | test error (%): 0.75 -> 0.66 | Helped |
Closest papers by meaning
- Variational Dropout Sparsifies Deep Neural Networks (ICML 2017 · similarity 0.90)
- Structured Dropout Variational Inference for Bayesian Neural Networks (NeurIPS 2021 · similarity 0.90)
- Variational Bayesian dropout: pitfalls and fixes (ICML 2018 · similarity 0.89)
- Dropout as a Structured Shrinkage Prior (ICML 2019 · similarity 0.88)
- Stochastic Modified Equations and Dynamics of Dropout Algorithm (ICLR 2024 · similarity 0.86)
- Sparse Variational Student-t Processes (AAAI 2024 · similarity 0.85)
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.