AI Paper Atlas

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

Sampling-Free Learning of Bayesian Quantized Neural Networks

Jiahao Su, Milan Cvitkovic, Furong Huang

ICLR 2020 · Probabilistic and scientific modeling › Uncertainty and Gaussian processes · approach: Approximate inference in Bayesian neural networks · 3 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 BQN instead of E-QNNMLP · MNISTNLL ($10^{-3}$): 546.6$ \pm $157.9 -> 130.0$\pm$3.5Improved
Use BQN instead of E-QNNMLP · KMNISTNLL ($10^{-3}$): 2385.6$\pm$432.3 -> 457.7$\pm$13.8Improved
Use BQN instead of E-QNNMLP · Fashion-MNISTNLL ($10^{-3}$): 2529.4$\pm$276.7 -> 417.3$\pm$8.1Improved
Use BQN instead of E-QNNCNN · MNISTNLL ($10^{-3}$): 425.3$\pm$61.8 -> 41.8$\pm$1.6Improved
Use BQN instead of E-QNNCNN · KMNISTNLL ($10^{-3}$): 3755.7$\pm$465.1 -> 295.5$\pm$1.4Improved
Use BQN instead of E-QNNCNN · Fashion-MNISTNLL ($10^{-3}$): 1610.7$\pm$158.4 -> 209.5$\pm$2.8Improved
Use BQN instead of E-QNNCNN · CIFAR10NLL ($10^{-3}$): 7989.7 $ \pm $ 600.2 -> 530.6 $ \pm $ 23.0Improved
Use BQN instead of E-QNNCNN · Fashion-MNIST% Err.: 3.02$\pm$0.37 -> 4.65$\pm$0.15Hurt
Use BQN instead of E-QNNCNN · MNIST% Err.: 0.85$\pm$0.13 -> 0.85$\pm$0.06No clear effect
Use sampling-free analytic inference rather than Monte Carlo sampling (ablation)BQN · MNIST, KMNIST, Fashion-MNISTaccuracy and NLLNo clear effect
Increase the regularization coefficient λ (ablation)BQN · CNN evaluationgap between analytic inference and MC samplingHurt
Remove the prior regularization term by setting λ = 0 (ablation)BQN · CNN evaluationposterior uncertaintyHurt

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