AI Paper Atlas

Research map / Probabilistic and scientific modeling / Variational and generative sampling

Locally Most Powerful Bayesian Test for Out-of-Distribution Detection using Deep Generative Models

Keunseo Kim, JunCheol Shin, Heeyoung Kim

NeurIPS 2021 · Probabilistic and scientific modeling › Variational and generative sampling · approach: Score-based generative modeling analysis · 5 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 LMPBT instead of input complexity (IC)VAE · CIFAR-10 -> SVHNAUROC: 0.912 -> 0.931Improved
use LMPBT instead of likelihood regret (LR)VAE · CIFAR-10 -> CelebAAUROC: 0.714 -> 0.783Improved
use LMPBT instead of likelihood regret (LR)VAE · CIFAR-100 -> CelebAAUROC: 0.598 -> 0.683Improved
use LMPBT instead of input complexity (IC)VAE · CIFAR-100 -> SVHNAUROC: 0.912 -> 0.956Improved
use LMPBT instead of likelihood ratio (LLR)VAE · CIFAR-10 -> MNISTAUROC: 0.008 -> 0.998Improved
use LMPBT instead of likelihood regret (LR)VAE · CIFAR-10 -> MNISTAUROC: 0.998 -> 0.998No clear effect
use top 30 Hessian eigenvalues instead of fewer than 10 (ablation)VAE · CIFAR-10 -> SVHNAUROC: 0.913Improved
increase number of Hessian eigenvalues beyond the optimum (ablation)VAE · CIFAR-10 -> SVHNAUROCMixed
use LMPBT instead of likelihood regret (LR) for costVAE · CIFAR-10seconds per sample: 0.30s -> 0.11sImproved
use LMPBT instead of LLR or IC for costVAE · CIFAR-10seconds per sample: 0.0125 s -> 0.11sHurt
scale Hessian eigendecomposition with parameters (ablation)VAE · FMNIST/CIFAR sizestime per parameter: 0.0028 s/dim -> 0.0025 s/dimNo clear effect
Hessian eigendecomposition cost relative to trainingVAE · FMNISTtime cost: 2543 s vs 13448 sNo clear effect

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