Research map / Probabilistic and scientific modeling / Uncertainty and Gaussian processes
Stationary Activations for Uncertainty Calibration in Deep Learning
Lassi Meronen, Christabella Irwanto, Arno Solin
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 |
|---|---|---|---|
| Use the Matérn-3/2 activation instead of ReLU | GoogLeNet with MC dropout · CIFAR-10 known classes, five classes held out during training | NLPD: 0.18 -> 0.11 | Improved |
| Use the Matérn activation instead of ReLU in the added classification layer | Radar emitter CNN · Radar emitter classification | uncertainty separation | Improved |
| Use Matérn-3/2 activations rather than the RBF activation (ablation) | UCI classifiers · Adult, Connect-4, Covtype, Diabetes | NLPD: Adult: $.327{\pm}.012 -> $.316{\pm}.007; Connect-4: $.473{\pm}.010 -> $.450{\pm}.008; Covtype: $.116{\pm}.002 -> $.118{\ | Mixed |
| Use a non-differentiable exponential activation instead of Matérn-3/2 (ablation) | UCI classifiers · Adult, Connect-4, Covtype, Diabetes | NLPD: Adult: $.316{\pm}.007 -> $.522{\pm}.084; Connect-4: $.450{\pm}.008 -> $.944{\pm}.056; Covtype: $.118{\pm}.003 -> $1.20{\ | Hurt |
| Use Matérn-5/2 instead of Matérn-3/2 (ablation) | UCI classifiers · Adult, Connect-4, Covtype, Diabetes | NLPD: Adult: $.316{\pm}.007 -> $.317{\pm}.007; Connect-4: $.450{\pm}.008 -> $.453{\pm}.009; Covtype: $.118{\pm}.003 -> $.115{\ | Mixed |
| Replace the Matérn-3/2 prior with an RBF prior (ablation) | UCI classification models · UCI classification benchmarks | NLPD | Hurt |
Closest papers by meaning
- Activation-level uncertainty in deep neural networks (ICLR 2021 · similarity 0.90)
- Periodic Activation Functions Induce Stationarity (NeurIPS 2021 · similarity 0.87)
- Linearization Turns Neural Operators into Function-Valued Gaussian Processes (ICML 2025 · similarity 0.87)
- Uncertainty Quantification with the Empirical Neural Tangent Kernel (NeurIPS 2025 · similarity 0.87)
- Exploring the Uncertainty Properties of Neural Networks’ Implicit Priors in the Infinite-Width Limit (ICLR 2021 · similarity 0.86)
- Bayesian Treatment of the Spectrum of the Empirical Kernel in (Sub)Linear-Width Neural Networks (ICLR 2025 · similarity 0.86)
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.
