TorchEBM Blog¶

Welcome to the TorchEBM blog! Here you'll find latest news, tutorials, research insights, and updates about the project.

Categories¶

Tutorials

Step-by-step guides to help you learn how to use TorchEBM effectively.

View Tutorials
Research

Cutting-edge research using energy-based models and insights from the field.

View Research
Announcements

Updates about TorchEBM releases, features, and roadmap.

View Announcements
Examples

Real-world examples and case studies using TorchEBM.

View Examples

Hamiltonian Mechanics

Pendulum phase space diagram

Hamiltonian mechanics is a way to describe how physical systems, like planets or pendulums, move over time, focusing on energy rather than just forces. By reframing complex dynamics through energy lenses, this 19th-century physics framework now powers cutting-edge generative AI. It uses generalized coordinates $q$ (like position) and their conjugate momenta $p$ (related to momentum), forming a phase space that captures the system's state. This approach is particularly useful for complex systems with many parts, making it easier to find patterns and conservation laws.

2024-04-05
in Tutorials, Examples
10 min read

Langevin Dynamics Sampling with TorchEBM

Langevin dynamics is a powerful sampling technique that allows us to draw samples from complex probability distributions. In this tutorial, we'll explore how to use TorchEBM's implementation of Langevin dynamics for sampling from various energy landscapes.