A Unified Risk View of Uncertainty: Posterior Risk for Disentanglement and Evaluation Beyond Proxies
cs.LG
Submitted: 2026-08-06
Updated: 2026-09-27
Code: https://github.com/laplax-org/laplax
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Terminology
Sources
- Understanding the bias-variance tradeoff of Bregman divergences
- Deep Evidential Regression
- An Axiomatic Assessment of Entropy- and Variance-based Uncertainty Quantification in Regression
- Disentangling Epistemic and Aleatoric Uncertainty in Reinforcement Learning
- FSP-Laplace: Function-Space Priors for the Laplace Approximation in Bayesian Deep Learning
- Uncertainty Quantification for Regression using Proper Scoring Rules
- How Reliable is Your Regression Model's Uncertainty Under Real-World Distribution Shifts?
- Bayesian Active Learning for Classification and Preference Learning
- Position: Epistemic uncertainty estimation methods are fundamentally incomplete
- Is Epistemic Uncertainty Faithfully Represented by Evidential Deep Learning Methods?
- Scale estimation and rate-unbiasedness for Gaussian processes under smoothness misspecification
- Maximum likelihood estimation and uncertainty quantification for Gaussian process approximation of deterministic functions
- A view on model misspecification in uncertainty quantification
- What Uncertainties Do We Need in Bayesian Deep Learning for Computer Vision?
- From Risk to Uncertainty: Generating Predictive Uncertainty Measures via Bayesian Estimation
- DEUP: Direct Epistemic Uncertainty Prediction
- Out-of-Distribution Detection Methods Answer the Wrong Questions
- A Generalized Bias-Variance Decomposition for Bregman Divergences
- ImageNet Large Scale Visual Recognition Challenge
- Quantification of Uncertainty with Adversarial Models
Related papers
- Polynomial-Augmented Neural Networks (PANNs) with Weak Orthogonality Constraints for Enhanced Function and PDE Approximation
- AIRL-S: Unifying Reinforcement Learning and Search-Based Test-Time Scaling via Adversarial Inverse Reinforcement Learning
- Transformers as Bayesian In-Context Experimenters: Smoothness-Adaptive Efficient ATE Estimation
- Convergence issues in Relational Concept Analysis based on AOC-posets
- Beliefs Beyond Posteriors: Local-Consistency Optimisation for Bayesian Neural Networks
- Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks