AI Paper Atlas

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

Accurate Layerwise Interpretable Competence Estimation

Vickram Rajendran, William LeVine

NeurIPS 2019 · Probabilistic and scientific modeling › Uncertainty and Gaussian processes · approach: Epistemic versus aleatoric decomposition · 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 ALICE instead of Trust Score for competence estimationSVM (RBF) · DIGITS, cross-entropymAP: .346 \textpm{}.086 -> .989 \textpm{} .0078Improved
Use ALICE instead of Trust Score for competence estimationMLP (5 iterations) · DIGITS, cross-entropymAP: .579 \textpm{} .12 -> .966 \textpm{} .047Improved
Use ALICE instead of Trust Score for competence estimationResNet50 (1 epoch) · CIFAR100, cross-entropymAP: .346 -> .723Improved
Add the in-distribution probability term p(D|x) (ablation)ALICE · CIFAR10/SVHN 10/90, distributional uncertaintymAP: .100 \raisebox{.2ex}{$\scriptstyle\pm$} 0.000 -> .868 \raisebox{.2ex}{$\scriptstyle\pm$} 0.014Improved
Add the conditional error indicator term p(E<δ|x,c_j) (ablation)ALICE · CIFAR100, cross-entropy, VGG16 underfitmAP: .0369 \textpm{} .0041 -> .913 \textpm{} .012Improved
Add the conditional error indicator term p(E<δ|x,c_j) (ablation)ALICE · DIGITS, cross-entropy, MLP underfitmAP: .0538 \textpm{} .031 -> .999 \textpm{} .0015Improved
Add the class-probability term p(c_j|x,D) (ablation)ALICE · Synthetic class overlap, 0.50 overlapmAP: .730 \textpm{} 0.0 -> .960 \textpm{} 0.0Improved
Use ALICE rather than Trust ScoreVGG16 (5 epochs) · CIFAR100, 0-1 errorAP: .762 -> .756Hurt
Use ALICE rather than Trust ScoreMLP (5 iterations) · DIGITS, 0-1 errorAP: .996 \textpm{} .0036 -> .984 \textpm{} .014Hurt

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