AI Paper Atlas

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

Explaining Deep Learning Models -- A Bayesian Non-parametric Approach

Wenbo Guo, Sui Huang, Yunzhe Tao, Xinyu Xing, Lin Lin

NeurIPS 2018 · Probabilistic and scientific modeling › Uncertainty and Gaussian processes · approach: Nonparametric and PAC-Bayes Bayesian learning · 30 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 DMM-MEN instead of LIMEExplaining MLP predictions · MNISTAccuracy: 99.95% -> 100%Improved
Use DMM-MEN instead of LIMEExplaining MLP predictions · MNISTProbability: 99.84% -> 99.89%Improved
Use DMM-MEN instead of LIMEExplaining CNN predictions · Fashion-MNISTAccuracy: 98.32% -> 100%Improved
Use DMM-MEN instead of LIMEExplaining CNN predictions · Fashion-MNISTProbability: 93.49% -> 97.59%Improved
Use DMM-MEN instead of LIMEExplaining VGG16 predictions · ImageNetAccuracy: 66.05% -> 85.6%Improved
Use DMM-MEN instead of LIMEExplaining VGG16 predictions · ImageNetProbability: 47.46% -> 69.36%Improved
Select top important pixels rather than random pixelsDMM-MEN global insights · MNIST and Fashion-MNISTPositive classification rateImproved
Construct cases from top 50 important pixels rather than randomly selected pixelsDMM-MEN global insights · MNIST and Fashion-MNISTPositive classification rate: around 50%Improved

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