This repository is a curated collection of research materials, including papers, open-source libraries, and tutorials, focused on Energy-Based Models (EBMs) and Energy-Based Learning (EBL). It serves as a comprehensive resource for researchers and practitioners interested in this area of machine learning, offering a structured overview of EBM applications and advancements.

How It Works

The repository organizes a vast amount of literature on EBMs, categorizing papers by publication year and providing links to code, talks, and workshops. It highlights the diverse applications of EBMs, ranging from data generation and density estimation to reinforcement learning and anomaly detection, showcasing their versatility across various domains.

Quick Start & Requirements

This repository is a curated list of resources and does not have a direct installation or execution command. Users are directed to individual papers and libraries for specific implementation details.

Highlighted Details

• Comprehensive reverse chronological listing of papers from 2025 back to early foundational work before 2007.
• Categorization of EBM applications including data generation, discriminative learning, OOD detection, and reinforcement learning.
• A list of open-source libraries like mini-ebm and TorchEBM for practical implementation.
• Links to key tutorials and talks from prominent researchers like Yann LeCun and Stefano Ermon.

Maintenance & Community

The repository is actively maintained, with recent updates and additions to papers and libraries. It serves as a community-driven effort to consolidate knowledge in the EBM field.

Licensing & Compatibility

The repository itself is a collection of links and does not have a specific license. Individual linked papers and libraries will have their own respective licenses.

Limitations & Caveats

Some entries, particularly for recent years (2024, 2025), are marked as "@todo" or "@to be updated," indicating that the collection is still growing and may not be fully comprehensive for the most current research. Some older code links may be broken or missing.