Research map / Label-efficient and robust learning / Recognition and long-tail
Uncertainty-Aware Attention for Reliable Interpretation and Prediction
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| replace deterministic attention with input-dependent Gaussian attention uncertainty | RETAIN · PhysioNet Mortality | AUROC: 0.7652$\pm$ 0.02 -> 0.7827$\pm$ 0.02 | Helped |
| replace deterministic attention with input-dependent Gaussian attention uncertainty | RETAIN · PhysioNet Stay < 3 | AUROC: 0.8515$\pm$ 0.02 -> 0.8628$\pm$ 0.02 | Helped |
| replace deterministic attention with input-dependent Gaussian attention uncertainty | RETAIN · PhysioNet Cardiac | AUROC: 0.9485$\pm$ 0.01 -> 0.9563$\pm$ 0.01 | Helped |
| replace deterministic attention with input-dependent Gaussian attention uncertainty | RETAIN · PhysioNet Recovery | AUROC: 0.8830$\pm$ 0.01 -> 0.9049$\pm$ 0.01 | Helped |
| replace deterministic attention with input-dependent Gaussian attention uncertainty | RETAIN · Pancreatic Cancer | AUROC: 0.8528$\pm$ 0.01 -> 0.8604$\pm$ 0.01 | Helped |
| replace deterministic attention with input-dependent Gaussian attention uncertainty | RETAIN · MIMIC Sepsis | AUROC: 0.7965$\pm$ 0.01 -> 0.8017$\pm$ 0.01 | Helped |
| make attention variance input-dependent rather than input-independent (ablation) | UA · PhysioNet Mortality | AUROC: 0.7764$\pm$ 0.01 -> 0.7827$\pm$ 0.02 | Helped |
| make attention variance input-dependent rather than input-independent (ablation) | UA · PhysioNet Stay < 3 | AUROC: 0.8572$\pm$ 0.02 -> 0.8628$\pm$ 0.02 | Helped |
| make attention variance input-dependent rather than input-independent (ablation) | UA · PhysioNet Cardiac | AUROC: 0.9516$\pm$ 0.01 -> 0.9563$\pm$ 0.01 | Helped |
| make attention variance input-dependent rather than input-independent (ablation) | UA · PhysioNet Recovery | AUROC: 0.8895$\pm$ 0.01 -> 0.9049$\pm$ 0.01 | Helped |
| make attention variance input-dependent rather than input-independent (ablation) | UA · Pancreatic Cancer | AUROC: 0.8533$\pm$ 0.03 -> 0.8604$\pm$ 0.01 | Helped |
| make attention variance input-dependent rather than input-independent (ablation) | UA · MIMIC Sepsis | AUROC: 0.8019$\pm$ 0.01 -> 0.8017$\pm$ 0.01 | No clear effect |
Closest papers by meaning
- Stop Guessing: Choosing the Optimization-Consistent Uncertainty Measurement for Evidential Deep Learning (ICLR 2026 · similarity 0.85)
- Enhancing Uncertainty Estimation and Interpretability with Bayesian Non-negative Decision Layer (ICLR 2025 · similarity 0.84)
- Are Uncertainty Quantification Capabilities of Evidential Deep Learning a Mirage? (NeurIPS 2024 · similarity 0.84)
- Uncertainty Quantification for Deep Regression using Contextualised Normalizing Flows (NeurIPS 2025 · similarity 0.83)
- Uncertainty Estimation by Flexible Evidential Deep Learning (NeurIPS 2025 · similarity 0.83)
- SURE: SUrvey REcipes for building reliable and robust deep networks (CVPR 2024 · similarity 0.83)
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.