Research map / Probabilistic and scientific modeling / Uncertainty and Gaussian processes
Gaussian Process Probes (GPP) for Uncertainty-Aware Probing
Zi Wang, Alexander Ku, Jason Michael Baldridge, Thomas L. Griffiths, Been Kim
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 GPP rather than bootstrap-ensembled linear probes (LPE) to estimate concept fuzziness | Probing 3D Shapes concepts with M.1 · 3D Shapes, P.1 concepts with injected label noise | Pearson correlation coefficient | Improved |
| Use GPP rather than LPE to estimate aleatory uncertainty for fuzzy concepts | Probing 3D Shapes concepts · 3D Shapes, ground truth label probability 0.5 | alea | Improved |
| Use GPP rather than linear probes (LP) or SVM probes for concept classification | Probing M.1, M.2, and M.3 · 3D Shapes, P.1 and P.2 | AUROC | No clear effect |
| Use GPP for out-of-distribution detection | Image representations from M.1, ResNet-50, and CoCa-Base · 3D Shapes and ImageNet-derived OOD tasks | AUROC | No clear effect |
Closest papers by meaning
- Posterior and Computational Uncertainty in Gaussian Processes (NeurIPS 2022 · similarity 0.86)
- Domain Invariant Learning for Gaussian Processes and Bayesian Exploration (AAAI 2024 · similarity 0.85)
- Computation-Aware Gaussian Processes: Model Selection And Linear-Time Inference (NeurIPS 2024 · similarity 0.85)
- Robust and Computation-Aware Gaussian Processes (NeurIPS 2025 · similarity 0.84)
- Conditioning Gaussian Processes on Almost Anything (NeurIPS 2026 · similarity 0.84)
- RAPTOR: Ridge-Adaptive Logistic Probes (NeurIPS 2026 · similarity 0.84)
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.
