AI Paper Atlas

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

Semantic uncertainty intervals for disentangled latent spaces

Swami Sankaranarayanan, Anastasios Nikolas Angelopoulos, Stephen Bates, Yaniv Romano, Phillip Isola

NeurIPS 2022 · Probabilistic and scientific modeling › Uncertainty and Gaussian processes · approach: Epistemic versus aleatoric decomposition · 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
Calibrate the latent quantile intervalsSemantic uncertainty intervals · Generated dataaverage risk: 0.114 -> 0.085Improved
Calibrate the latent quantile intervalsSemantic uncertainty intervals · Real dataaverage risk: 0.267 -> 0.096Improved
Increase input corruption from Easy to HardSemantic uncertainty intervals · Generated and real dataaverage set size: Generated data: 2.5 -> 4.7; Real data: 3.1 -> 5.3Mixed
Increase corruption difficulty from Easy to HardSemantic uncertainty intervals on FFHQ · FFHQID Loss between interval endpoints: 0.03 -> 0.08Improved
Increase corruption difficulty from Easy to HardSemantic uncertainty intervals on FFHQ · FFHQPerceptual Loss between interval endpoints: 0.17 -> 0.24Improved
Set feature coefficient c to 0 instead of 10.0 (ablation)Semantic uncertainty intervals · Easy corruptionpredicted set size: 2.5 -> 2.56Mixed
Set feature coefficient c to 1.0 instead of 10.0 (ablation)Semantic uncertainty intervals · Easy corruptionpredicted set size: 2.5 -> 2.4Mixed
Set feature coefficient c to 0 instead of 10.0 (ablation)Semantic uncertainty intervals · Hard corruptionpredicted set size: 4.7 -> 5.23Hurt
Set feature coefficient c to 1.0 instead of 10.0 (ablation)Semantic uncertainty intervals · Hard corruptionpredicted set size: 4.7 -> 5.1Hurt

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