Research map / Probabilistic and scientific modeling / Variational and generative sampling
Measuring Sample Quality in Algorithms for Intractable Normalizing Function Problems
Bokgyeong Kang, John Hughes, Murali Haran
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 IMQ KSD instead of curvature diagnostic | multivariate normality test · p=25, n=1000 | Type I error rate: 0.13 -> 0.00 | Improved |
| increase sample size from 1000 to 5000 | curvature diagnostic · p=25 | Type I error rate: 0.13 -> 0.02 | Improved |
| compare CD and KSD power | multivariate normality test · n=1000..5000, p=2..25 | power: 1 for both | No clear effect |
| select d by ACD/AIKS instead of ESS | AEX on Ising model · Ising 30x30, theta=0.2 | tuning choice: d=100 -> d=470 | Improved |
| use ACD/AIKS-recommended m=4 instead of ESS-best m=1 | DMH on Ising model · Ising 30x30 | right-tail probability: 0.09 -> 0.05 | Improved |
| use truncation k=10 instead of ESS-best k=3 | NormTrunc on COM-Poisson · COM-Poisson regression beta_0 | median: 0.45 -> 0.31 | Improved |
| use truncation k=10 instead of k=3 | NormTrunc on COM-Poisson · COM-Poisson regression beta_0 | right-tail probability: 0.47 -> 0.05 | Improved |
| add more particles d for ALR | ALR on Ising model · Ising 30x30 | diagnostics | No clear effect |
| use ACD instead of AIKS for large n | diagnostic cost · Ising model | computation time: 27 seconds -> 3 seconds | Improved |
Closest papers by meaning
- Complexity Analysis of Normalizing Constant Estimation: from Jarzynski Equality to Annealed Importance Sampling and beyond (ICLR 2026 · similarity 0.87)
- Kernelized Normalizing Constant Estimation: Bridging Bayesian Quadrature and Bayesian Optimization (AAAI 2024 · similarity 0.86)
- Multi-Marginal Couplings for Metropolis--Hastings (NeurIPS 2026 · similarity 0.86)
- Constrained Sampling with Primal-Dual Langevin Monte Carlo (NeurIPS 2024 · similarity 0.85)
- The Polynomial Stein Discrepancy for Assessing Moment Convergence (ICML 2025 · similarity 0.84)
- Provable Posterior Sampling with Denoising Oracles via Tilted Transport (NeurIPS 2024 · 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.
