This repository serves as a curated bibliography of research papers focused on Flow Matching and Stochastic Interpolants, primarily within the context of generative modeling. It aims to provide a comprehensive overview of the latest advancements, theoretical underpinnings, and diverse applications of these techniques in machine learning, acting as a valuable resource for researchers and practitioners.

How It Works

This repository functions as a curated bibliography, compiling and organizing academic publications related to flow matching and stochastic interpolants. It systematically lists papers, their authors, and publication venues, providing a structured overview of research in this domain. The collection aims to serve as a comprehensive resource for understanding the evolution and application of these generative modeling techniques.

Quick Start & Requirements

This section is not applicable as the repository is a list of research papers and not a software project with installation instructions.

Highlighted Details

• Features a broad spectrum of research from top-tier conferences (ICLR, ICML, NeurIPS) and arXiv, with publications spanning 2020 to 2025.
• Encompasses diverse applications, including image synthesis, speech generation, protein design, physics-constrained modeling, and autonomous driving.
• Highlights influential works such as "Flow Matching Guide and Code" (152 citations), "FLUX.1 Kontext" (368 citations), and "F5-TTS" (285 citations), indicating significant community impact.

Maintenance & Community

The list is automatically generated and updated daily, suggesting a dynamic and continuously expanding resource. Specific community channels or direct contributor information are not detailed within this README.

Licensing & Compatibility

No explicit license is provided for the repository itself. As a compilation of research papers, usage of the information within each paper would be governed by the respective publication's license and copyright. Compatibility for commercial use or closed-source linking would depend on the individual paper licenses.

Limitations & Caveats

This repository is a literature index and does not provide code, implementations, or direct tutorials. Researchers and practitioners must consult the individual papers for practical application. The automated generation process may lead to varying levels of relevance or curation for some entries.