AI Paper Atlas

Research map / Probabilistic and scientific modeling / Uncertainty and Gaussian processes

Should We Learn Most Likely Functions or Parameters?

Shikai Qiu, Tim G. J. Rudner, Sanyam Kapoor, Andrew Gordon Wilson

NeurIPS 2023 · Probabilistic and scientific modeling › Uncertainty and Gaussian processes · approach: Simulation-based inference · 0 citations (OpenAlex)

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 changedWhereMeasuredEffect
function-space MAP (Laplacian approximation) instead of parameter-space MAPMLP 3x256 · UCI regression (8 datasets)normalized test RMSE: .032 -> .018Improved
L-MAP instead of PS-MAPMLP 3x256 · UCI Bostonnormalized test RMSE: .329 -> .352Hurt
L-MAP instead of PS-MAPResNet-18 · CIFAR-10accuracy: 95.4 -> 95.3No clear effect
L-MAP with CIFAR-100 evaluation points instead of PS-MAPResNet-18 · CIFAR-10ECE: 2.5 -> 1.4Improved
L-MAP instead of PS-MAP in ImageNet transferResNet-50 · CIFAR-10NLL: 0.18 -> 0.15Improved
L-MAP instead of PS-MAP in ImageNet transferResNet-50 · CIFAR-10ECE: 2.6 -> 2.1Improved
L-MAP with KMNIST evaluation points instead of PS-MAPResNet-18 · FashionMNISTaccuracy: 93.9 -> 94.1No clear effect
use external evaluation set (CIFAR-100) rather than N(0,I) noise for evaluation distribution (ablation)L-MAP · CIFAR-10NLL: .20 -> .16Improved
increase Monte Carlo samples S for Laplacian estimator (ablation)L-MAP · CIFAR-10accuracyNo clear effect
increase Monte Carlo samples S for Laplacian estimator (ablation)L-MAP · CIFAR-10calibration errorImproved
training with label noise: L-MAP vs PS-MAPL-MAP · CIFAR-10test performance under label noiseImproved
increase number of evaluation points up to 200 (ablation)FS-MAP · synthetic Fourier regression, 400 training samplestest RMSEImproved

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy