Research map / Probabilistic and scientific modeling / Variational and generative sampling
Sampling-Based Accuracy Testing of Posterior Estimators for General Inference
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 TARP coverage with reference points depending on the observation, rather than HPD coverage | Posterior accuracy testing · Biased Gaussian toy model | Expected coverage probability | Helped |
| Use TARP coverage with observation-dependent reference points | Posterior accuracy testing · Uninformative estimator equal to the prior | Expected coverage probability | Helped |
| Use reference points independent of the observation (ablation) | TARP · Uninformative estimator equal to the prior | Expected coverage probability | No clear effect |
| Use HPD regions instead of positionable TARP regions (ablation) | Posterior accuracy testing · Biased Gaussian toy model | Expected coverage probability | No clear effect |
| Use L1 rather than L2 distance (ablation) | TARP · Gaussian toy model | Expected coverage probability | No clear effect |
| Vary the reference-point distribution (ablation) | TARP · Gaussian toy model | Expected coverage probability | No clear effect |
| Add non-zero off-diagonal covariance elements (ablation) | Gaussian toy model · Gaussian toy model | Expected coverage probability | No clear effect |
| Apply TARP to a high-dimensional posterior | TARP · Gaussian toy model, D = 1000 | Expected coverage probability | Helped |
| Use the exact rather than biased posterior sampler (ablation) | Gravitational lensing posterior inference · Gravitational lensing | Expected coverage probability | Helped |
Closest papers by meaning
- Addressing Misspecification in Simulation-based Inference through Data-driven Calibration (ICML 2025 · similarity 0.85)
- A Black-Box Debiasing Framework for Conditional Sampling (NeurIPS 2025 · similarity 0.85)
- Consistency Models for Scalable and Fast Simulation-Based Inference (NeurIPS 2024 · similarity 0.85)
- Understanding the difficulties of posterior predictive estimation (ICML 2025 · similarity 0.84)
- The Limits of Inference Scaling Through Resampling (ICLR 2026 · similarity 0.84)
- Experts Don't Cheat: Learning What You Don't Know By Predicting Pairs (ICML 2024 · 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.