AI Paper Atlas

Research map / Probabilistic and scientific modeling / Uncertainty and Gaussian processes

Inverse-Reference Priors for Fisher Regularization of Bayesian Neural Networks

Keunseo Kim, Eun-Yeol Ma, Jeongman Choi, Heeyoung Kim

AAAI 2023 · Probabilistic and scientific modeling › Uncertainty and Gaussian processes · approach: Approximate inference in Bayesian neural networks · 1 citations (OpenAlex)

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 changedWhereMeasuredEffect
Use the IR prior instead of the strongest competing priorDenseNet121 BNN · CIFAR-10validation accuracy (%): 81.9 -> 85.5Improved
Use the IR prior instead of the strongest competing priorResNet18 BNN · CIFAR-10validation accuracy (%): 82.6 -> 84.6Improved
Use the IR prior instead of the strongest competing priorVGG16 BNN · CIFAR-10validation accuracy (%): 79.4 -> 84.1Improved
Use the IR prior instead of the strongest competing priorDenseNet121 BNN · CIFAR-100validation accuracy (%): 48.3 -> 52.0Improved
Use the IR prior instead of the strongest competing priorResNet18 BNN · CIFAR-100validation accuracy (%): 48.9 -> 51.1Improved
Use the IR prior instead of the strongest competing priorVGG16 BNN · CIFAR-100validation accuracy (%): 48.1 -> 50.8Improved
Use the IR prior instead of the strongest competing priorDenseNet121 BNN · SVHNvalidation accuracy (%): 94.9 -> 96.1Improved
Use the IR prior instead of the strongest competing priorResNet18 BNN · SVHNvalidation accuracy (%): 94.8 -> 95.4Improved
Use the IR prior instead of the strongest competing priorVGG16 BNN · SVHNvalidation accuracy (%): 94.3 -> 95.4Improved
Use the IR prior instead of the strongest competing priorDenseNet BNN · CIFAR-10 adversarial examples (ϵ = 0.0001)validation accuracy (%): 80.1 -> 84.9Improved
Use the IR prior instead of the strongest competing priorDenseNet BNN · CIFAR-10 adversarial examples (ϵ = 0.001)validation accuracy (%): 78.9 -> 84.7Improved
Use the IR prior instead of the strongest competing priorDenseNet BNN · CIFAR-100 adversarial examples (ϵ = 0.0001)validation accuracy (%): 47.6 -> 50.2Improved

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy