Research map / Probabilistic and scientific modeling / Uncertainty and Gaussian processes
Regression uncertainty quantification
Estimates predictive uncertainty for deep and high-dimensional regression.
43 accepted papers at ICML, NeurIPS, ICLR, CVPR and AAAI (2016–2026) · an approach in Uncertainty and Gaussian processes, keyword cluster “conformal · intervals · prediction”
State History Tried Checklist Preprints Who Export
Where it stands
Not enough benchmark data Fewer than four papers report the same benchmark, so there is no record to follow.
11 papers in 2023–24 and 17 in 2025–26 (at ICML, NeurIPS, ICLR, CVPR and AAAI); its share of the topic grew, from 4.0% to 5.5%.
How this is decided
A benchmark is tracked when at least four papers of this approach report their own method on the same dataset and metric in at least two different years. Its record is the best number reported up to 2024 compared with the best reported in 2025–26: still improving means at least 1% better, flat means at most 0.2% better with at least two papers from 2025–26 trying. Numbers on a different scale from the rest (a fraction among percentages) and far outliers are left out. Results come from different papers with different setups, so they are signals, not head-to-head comparisons. The share of papers is counted at ICML, NeurIPS, ICLR, CVPR and AAAI.
How it got here
Most cited paper of each year
- 2016 · Testing for Differences in Gaussian Graphical Models: Applications to Brain Connectivity (NeurIPS 2016 · 21 citations)
- 2018 · High-Quality Prediction Intervals for Deep Learning: A Distribution-Free, Ensembled Approach (ICML 2018 · 150 citations)
- 2019 · Learning for Single-Shot Confidence Calibration in Deep Neural Networks Through Stochastic Inferences (CVPR 2019 · 71 citations)
- 2020 · Discriminative Jackknife: Quantifying Uncertainty in Deep Learning via Higher-Order Influence Functions (ICML 2020 · 13 citations)
- 2021 · Beyond Pinball Loss: Quantile Methods for Calibrated Uncertainty Quantification (NeurIPS 2021 · 15 citations)
- 2022 · Training Uncertainty-Aware Classifiers with Conformalized Deep Learning (NeurIPS 2022 · 17 citations)
- 2023 · Efficient Uncertainty Quantification and Reduction for Over-Parameterized Neural Networks (NeurIPS 2023 · 2 citations)
- 2024 · Robust Uncertainty Quantification Using Conformalised Monte Carlo Prediction (AAAI 2024 · 23 citations)
- 2025 · Combining Priors with Experience: Confidence Calibration Based on Binomial Process Modeling (AAAI 2025 · 4 citations)
- 2026 · Beta Distribution Learning for Reliable Roadway Crash Risk Assessment (AAAI 2026 · 3 citations)
Earliest papers here
- Testing for Differences in Gaussian Graphical Models: Applications to Brain Connectivity (NeurIPS 2016 · 21 citations)
- High-Quality Prediction Intervals for Deep Learning: A Distribution-Free, Ensembled Approach (ICML 2018 · 150 citations)
- Learning for Single-Shot Confidence Calibration in Deep Neural Networks Through Stochastic Inferences (CVPR 2019 · 71 citations)
What has been tried
34 papers here report 296 measured changes (a component added, removed or swapped, compared with a baseline): helped 62%, no clear effect 7%, hurt 24%, mixed 6%. Papers mostly publish changes that helped, so "helped" is over-represented.
Ablations in recent papers
| What was changed | Measured on | Result | Effect | Paper |
|---|---|---|---|---|
| Use δ=0.1 and P=1 rather than δ=0.00001 and P=100 | CIFAR-10 · test error | 1.50 ± 0.48 -> 6.25 ± 0.43 | Hurt | Robust Uncertainty Quantification Using Conformalised Monte Carlo Prediction (AAAI 2024 · 23 citations) |
| Add noise-scaling reparametrization of the score function | Probabilistic prediction on real-world datasets · accuracy and stability | Helped | Treeffuser: probabilistic prediction via conditional diffusions with gradient-boosted trees (NeurIPS 2024 · 2 citations) | |
| Use the Gaussian kernel rather than the Laplace kernel | Targeted GP sampling · information from points outside the target space | Mixed | Transductive Active Learning: Theory and Applications (NeurIPS 2024 · 1 citation) | |
| Add MCD prediction-error resampling to propagate epistemic uncertainty | Protein · median MAE | 0.186±0.008 -> 0.202±0.009 | Helped | Uncertainty Quantification for Deep Regression using Contextualised Normalizing Flows (NeurIPS 2025 · 0 citations) |
| Set dropout rate p=0 | Asymptotic relative efficiency vs. vanilla Boulevard · ARE | ∈ [1,4] | Hurt | Statistical Inference for Gradient Boosting Regression (NeurIPS 2025 · 0 citations) |
| Vary L from 0.1 to 10 | Single covariate simulation · coverage | coverage of 1.0 for all L values | No clear effect | Smooth Sailing: Lipschitz-Driven Uncertainty Quantification for Spatial Associations (NeurIPS 2025 · 0 citations) |
| use a p-dependent IQT update instead of a p-independent update | Amazon stock prices, AR, infrequent observations · prediction interval size | Mixed | Conformalized Interactive Imitation Learning: Handling Expert Shift and Intermittent Feedback (ICLR 2025 · 0 citations) |
What reviewers will expect
Datasets and baselines named in the 11 papers of this approach from 2025–2026 that have a full-text summary. Leaving out the common ones is a frequent review question.
Datasets
- Fashion-MNIST · 2 of 11 papers
- CIFAR-10 · 2 of 11 papers
- CIFAR-100 · 2 of 11 papers
Baselines
- XGBoost · 2 of 11 papers
Recent preprints
Loading recent arXiv preprints…
Who publishes on it
Institutions on the most papers of this approach since 2023 (of 28 papers; from OpenAlex and the papers' affiliations).
- University of York · 2 papers
- Stanford University · 2 papers
- Columbia University · 2 papers
- Ludwig-Maximilians-Universität München · 2 papers
- Harvard University · 2 papers
Use it
Download BibTeX of 35 key papersOpen the topic in the interactive map
Get alerts. Describe your idea in scime atlas and save it (free scime account): you get a weekly email when new preprints land close to it.
Related approaches
Same keyword cluster
- Conformal prediction methods (29 papers)
Elsewhere in the topic
- Nonparametric and PAC-Bayes Bayesian learning (130 papers)
- Simulation-based inference (118 papers)
- Priors and functional variational uncertainty (112 papers)
- Variational Bayesian deep learning (105 papers)
- Approximate inference in Bayesian neural networks (93 papers)
- Sparse variational Gaussian processes (92 papers)
- PAC-Bayes and decoupled GP learning (83 papers)
- Epistemic versus aleatoric decomposition (75 papers)
Benchmark numbers, ablations and baseline numbers are extracted by AI from each paper's full text and can be wrong; the paper is the authority. Data: results index of 2026-11-02, citation counts from OpenAlex (2026-10-09).