Research map / Probabilistic and scientific modeling / Variational and generative sampling
Stein Transport for Generative Modeling
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's abstract only. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Proposes a nonparametric, kernel-based generative model that transports particles from a simple reference (e.g., standard Gaussian) to a target distribution via a Stein formulation of the continuity equation.
Problem
Generative modeling with transport maps that avoids parametric function approximation; existing Stein transport methods require access to the target density or its gradient.
Approach
Builds a continuous-time transport from a reference distribution to the target using a Stein formulation of the continuity equation, in a nonparametric kernel framework; unlike prior Stein transport methods, it does not require target density or gradient access (abstract is truncated, so the exact mechanism is not fully stated).
Experimental setup
Builds on: Stein variational gradient descent
Closest papers by meaning
- Analysis of Learning a Flow-based Generative Model from Limited Sample Complexity (ICLR 2024 · similarity 0.87)
- Generative Conditional Distributions by Neural (Entropic) Optimal Transport (ICML 2024 · similarity 0.87)
- Multimarginal Generative Modeling with Stochastic Interpolants (ICLR 2024 · similarity 0.87)
- Towards Non-Asymptotic Convergence for Diffusion-Based Generative Models (ICLR 2024 · similarity 0.86)
- Generative Modeling by Value-Driven Transport (NeurIPS 2026 · similarity 0.86)
- Statistically Optimal Generative Modeling with Maximum Deviation from the Empirical Distribution (ICML 2024 · 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.