Research map / Probabilistic and scientific modeling / Uncertainty and Gaussian processes
Martingale Posterior Neural 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 |
|---|---|---|---|
| Replace a fixed-dimensional latent uncertainty representation with martingale-posterior uncertainty from generated pseudo contexts | Neural Processes · 1D regression, RBF; image completion on MNIST, SVHN, and CelebA | target or task log-likelihood: 0.515 -> 0.675; 0.706 -> 0.757; 2.796 -> 2.992; 2.134 -> 2.407 | Helped |
| Replace attention NP uncertainty with martingale-posterior uncertainty from generated pseudo contexts | Attentive Neural Processes · 1D regression, RBF; image completion on MNIST, SVHN, and CelebA | target or task log-likelihood: 0.850 -> 0.881; 0.793 -> 0.862; 3.405 -> 3.544; 2.833 -> 2.934 | Helped |
| Use martingale-posterior uncertainty from generated pseudo contexts | MPNP · Predator-prey simulated data | task log-likelihood: 0.417 -> 0.500 | Helped |
| Use martingale-posterior uncertainty from generated pseudo contexts | MPNP · Predator-prey real data | task log-likelihood: -2.868 -> -2.918 | Hurt |
| Generate pseudo-context representations rather than raw input data | MPNP and MPANP · Regression and image completion tasks | predictive performance | Helped |
| Generate pseudo-context data directly in input space | MPNP · Direct input generation experiments | predictive performance | Hurt |
| Increase baseline encoder size to match the proposed model's parameter count | CNP, NP, BNP, CANP, ANP, and BANP · 1D regression, RBF and Matern | target log-likelihood | No clear effect |
Closest papers by meaning
- Martingale Posterior Neural Networks for Fast Sequential Decision Making (NeurIPS 2025 · similarity 0.86)
- Rényi Neural Processes (ICML 2025 · similarity 0.85)
- Flow Matching Neural Processes (NeurIPS 2025 · similarity 0.85)
- Efficient Training of Neural Stochastic Differential Equations by Matching Finite Dimensional Distributions (ICLR 2025 · similarity 0.84)
- Distance-informed Neural Processes (NeurIPS 2025 · similarity 0.83)
- Quantifying Uncertainty in the Presence of Distribution Shifts (NeurIPS 2025 · similarity 0.83)
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.