Research map / Probabilistic and scientific modeling / Variational and generative sampling
Learning Optimal Flows for Non-Equilibrium Importance Sampling
Yu Cao, Eric Vanden-Eijnden
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 |
|---|---|---|---|
| train flow velocity by minimizing estimator variance with assisted SGD | NEIS · 2D asymmetric 2-mode Gaussian mixture | variance of Z1 estimator: 10^6 -> 1 | Improved |
| train flow velocity by minimizing estimator variance (gradient ansatz, 60 SGD steps) | NEIS · 10D symmetric 4-mode Gaussian mixture | variance of Z1 estimator: 2.15\times 10^6 -> 10 | Improved |
| NEIS with optimized flow instead of AIS-100 under equal query budget | NEIS · 2D Gaussian mixture | std of Z1 estimate | Improved |
| NEIS with optimized flow instead of AIS-100 under equal query budget | NEIS · 10D Gaussian mixture | std of Z1 estimate | Improved |
| use two-parametric ansatz instead of generic linear ansatz for velocity field (ablation) | NEIS · 10D funnel distribution | Z1 estimation accuracy | Improved |
| use generic linear ansatz NEIS instead of AIS-100 | NEIS · 10D funnel distribution | Z1 estimation accuracy | Improved |
| use asymmetric t_-=0 instead of symmetric t_-=-1/2 (ablation) | NEIS · 10D funnel distribution | Z1 estimation accuracy | Hurt |
| bias from ODE time discretization | NEIS · Gaussian mixtures | bias of Z1 estimate | No clear effect |
Closest papers by meaning
- Liouville Flow Importance Sampler (ICML 2024 · similarity 0.90)
- NETS: A Non-equilibrium Transport Sampler (ICML 2025 · similarity 0.89)
- Fast and unified path gradient estimators for normalizing flows (ICLR 2024 · similarity 0.87)
- Importance Corrected Neural JKO Sampling (ICML 2025 · similarity 0.86)
- Enhanced Importance Sampling Through Latent Space Exploration in Normalizing Flows (AAAI 2025 · similarity 0.86)
- Annealing Flow Generative Models Towards Sampling High-Dimensional and Multi-Modal Distributions (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.
