Research map / Probabilistic and scientific modeling / Uncertainty and Gaussian processes
Function-Space Distributions over Kernels
Gregory Benton, Wesley J Maddox, Jayson Salkey, Julio Albinati, Andrew Gordon Wilson
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 |
|---|---|---|---|
| share one latent GP across input dimensions instead of separate latent GPs (ablation) | FKL · UCI concreteslump | RMSE: 59.444 -> 4.385 | Improved |
| share one latent GP across input dimensions instead of separate latent GPs (ablation) | FKL · UCI housing | RMSE: 4.679 -> 2.703 | Improved |
| share one latent GP across input dimensions instead of separate latent GPs (ablation) | FKL · UCI yacht | RMSE: 0.193 -> 0.11 | Improved |
| share one latent GP across input dimensions instead of separate latent GPs (ablation) | FKL · UCI energy | RMSE: 0.296 -> 0.334 | No clear effect |
| use FKL shared instead of ARD Matern kernel | FKL · UCI concreteslump | RMSE: 26.335 -> 4.385 | Improved |
| use FKL shared instead of ARD Matern kernel | FKL · UCI concrete | RMSE: 6.058 -> 4.047 | Improved |
| use FKL shared instead of ARD Matern kernel | FKL · UCI stock | RMSE: 0.005 -> 0.016 | Hurt |
| use FKL shared instead of ARD Matern kernel | FKL · UCI pendulum | RMSE: 1.013 -> 1.562 | Hurt |
| use FKL shared instead of ARD Matern kernel | FKL · UCI concreteslump | standardized log loss: 17.247 -> -2.57 | Improved |
| use FKL shared instead of ARD Matern kernel | FKL · UCI airfoil | negative log-likelihood: 305.588 -> 270.073 | Improved |
| use multi-task FKL instead of collaborative GP (CGP) | FKL · FX exchange rates | SMSE: 0.2125 -> 0.1392 | Improved |
| use GPAR instead of multi-task FKL | FKL · FX exchange rates | SMSE: 0.1392 -> 0.0302 | Improved |
Closest papers by meaning
- Bayesian Deconditional Kernel Mean Embeddings (ICML 2019 · similarity 0.89)
- Scalable Levy Process Priors for Spectral Kernel Learning (NeurIPS 2017 · similarity 0.88)
- Non-Stationary Spectral Kernels (NeurIPS 2017 · similarity 0.88)
- Scalable Log Determinants for Gaussian Process Kernel Learning (NeurIPS 2017 · similarity 0.87)
- Revisiting Nonstationary Kernel Design for Multi-Output Gaussian Processes (ICLR 2026 · similarity 0.87)
- Automated Kernel Discovery Towards Understanding High-dimensional Bayesian Optimization (NeurIPS 2026 · similarity 0.87)
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.
