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
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 changed | Where | Measured | Effect |
|---|---|---|---|
| Calibrate the latent quantile intervals | Semantic uncertainty intervals · Generated data | average risk: 0.114 -> 0.085 | Improved |
| Calibrate the latent quantile intervals | Semantic uncertainty intervals · Real data | average risk: 0.267 -> 0.096 | Improved |
| Increase input corruption from Easy to Hard | Semantic uncertainty intervals · Generated and real data | average set size: Generated data: 2.5 -> 4.7; Real data: 3.1 -> 5.3 | Mixed |
| Increase corruption difficulty from Easy to Hard | Semantic uncertainty intervals on FFHQ · FFHQ | ID Loss between interval endpoints: 0.03 -> 0.08 | Improved |
| Increase corruption difficulty from Easy to Hard | Semantic uncertainty intervals on FFHQ · FFHQ | Perceptual Loss between interval endpoints: 0.17 -> 0.24 | Improved |
| Set feature coefficient c to 0 instead of 10.0 (ablation) | Semantic uncertainty intervals · Easy corruption | predicted set size: 2.5 -> 2.56 | Mixed |
| Set feature coefficient c to 1.0 instead of 10.0 (ablation) | Semantic uncertainty intervals · Easy corruption | predicted set size: 2.5 -> 2.4 | Mixed |
| Set feature coefficient c to 0 instead of 10.0 (ablation) | Semantic uncertainty intervals · Hard corruption | predicted set size: 4.7 -> 5.23 | Hurt |
| Set feature coefficient c to 1.0 instead of 10.0 (ablation) | Semantic uncertainty intervals · Hard corruption | predicted set size: 4.7 -> 5.1 | Hurt |
Closest papers by meaning
- Benchmarking Uncertainty Disentanglement: Specialized Uncertainties for Specialized Tasks (NeurIPS 2024 · similarity 0.88)
- UncertainSAM: Fast and Efficient Uncertainty Quantification of the Segment Anything Model (ICML 2025 · similarity 0.87)
- ValUES: A Framework for Systematic Validation of Uncertainty Estimation in Semantic Segmentation (ICLR 2024 · similarity 0.87)
- Towards Understanding and Quantifying Uncertainty for Text-to-Image Generation (CVPR 2025 · similarity 0.86)
- Epistemic Uncertainty for Generated Image Detection (NeurIPS 2025 · similarity 0.85)
- FUSE: Quantifying Uncertainty in Vision-Language Models by Bayesian Fusing Epistemic and Aleatoric Uncertainty (ICML 2026 · similarity 0.85)
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.
