Research map / Probabilistic and scientific modeling / Variational and generative sampling
Dynamic Importance Sampling for Anytime Bounds of the Partition Function
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 |
|---|---|---|---|
| interleave importance sampling with AOBFS search (DIS, Nl=1, Nd=10) | WMB-IS sampling and AOBFS search · pedigree, protein, BN benchmarks | normalized mean area between logZ bounds (lower better): pedigree 0.585 vs 1; protein 0.095 vs 1; BN 0.162 vs 1 | Helped |
| two-stage sampling (search to memory limit, then sample) instead of interleaving (ablation) | DIS · protein benchmark | normalized mean area (lower better): 2.511 vs 0.095 for DIS (Nl=1, Nd=10) | Hurt |
| two-stage sampling instead of interleaving (ablation) | DIS · pedigree benchmark | normalized mean area (lower better): 1.321 vs 0.585 for DIS (Nl=1, Nd=10) | Hurt |
| increase search expansions per sample from Nd=1 to Nd=10 (ablation) | DIS · BN benchmark | normalized mean area (lower better): 0.340 to 0.162 | Helped |
| increase search expansions per sample from Nd=1 to Nd=10 (ablation) | DIS · pedigree and protein benchmarks | normalized mean area (lower better): pedigree 0.711 to 0.585; protein 0.110 to 0.095 | Helped |
| DIS relative to pure AOBFS search | AOBFS · pedigree benchmark | normalized mean area (lower better): 16.638 for AOBFS vs 0.585 for DIS | Helped |
| DIS relative to pure WMB-IS sampling | WMB-IS · protein 1bgc and 1co6 instances | anytime bound quality | Helped |
Closest papers by meaning
- Anytime Anyspace AND/OR Search for Bounding the Partition Function (AAAI 2017 · similarity 0.91)
- Interleave Variational Optimization with Monte Carlo Sampling: A Tale of Two Approximate Inference Paradigms (AAAI 2019 · similarity 0.90)
- Stochastic Discrete Clenshaw-Curtis Quadrature (ICML 2016 · similarity 0.88)
- Beyond ELBOs: A Large-Scale Evaluation of Variational Methods for Sampling (ICML 2024 · similarity 0.88)
- Anytime Anyspace AND/OR Best-First Search for Bounding Marginal MAP (AAAI 2018 · similarity 0.87)
- A Unified Framework for Entropy Search and Expected Improvement in Bayesian Optimization (ICML 2025 · similarity 0.86)
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.