Research map / Probabilistic and scientific modeling / Uncertainty and Gaussian processes
Function-space Inference with Sparse Implicit Processes
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 SIP with a neural sampler prior instead of VIP | UCI multivariate regression · UCI datasets (average rank across datasets and splits) | RMSE rank: 3.29$ \pm $0.09 -> 3.10$ \pm $0.09 | Helped |
| Use SIP with a BNN prior instead of AADM | UCI multivariate regression · UCI datasets (average rank across datasets and splits) | NLL rank: 2.82$ \pm $0.06 -> 2.76$ \pm $0.08 | Helped |
| Use SIP with a BNN prior instead of VIP | UCI multivariate regression · UCI datasets (average rank across datasets and splits) | CRPS rank: 2.57$ \pm $0.08 -> 2.45$ \pm $0.08 | Helped |
| Replace the symmetrized KL objective with only the forward KL term (ablation) | SIP_BNN · Bimodal synthetic regression | prior adaptation | Hurt |
| Add VIP's regularization term (ablation) | VIP · Bimodal synthetic regression | predictive-distribution flexibility | Hurt |
| Add VIP's regularization term (ablation) | VIP · Heteroscedastic synthetic regression | heteroscedasticity capture | Hurt |
| Use adaptive contrast for the discriminator-based KL estimate (ablation) | SIP · Regression experiments | prior training | Hurt |
| Increase the number of inducing points beyond 50–100 (ablation) | SIP · Regression experiments | accuracy | Helped |
Closest papers by meaning
- Deep Variational Implicit Processes (ICLR 2023 · similarity 0.91)
- Input Dependent Sparse Gaussian Processes (ICML 2022 · similarity 0.87)
- Turbocharging Gaussian Process Inference with Approximate Sketch-and-Project (NeurIPS 2025 · similarity 0.86)
- Sparse Variational Student-t Processes (AAAI 2024 · similarity 0.85)
- Sparse Gaussian Processes: Structured Approximations and Power-EP Revisited (NeurIPS 2025 · similarity 0.85)
- New Bounds for Sparse Variational Gaussian Processes (ICML 2025 · similarity 0.85)
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.