Equilibrium Matching: Generative Modeling with Implicit Energy-Based Models
cs.LG, cs.AI, cs.CV
Submitted: 2025-10-02
Updated: 2026-11-06
Project page: https://raywang4.github.io/equilibrium_matching
License: http://creativecommons.org/licenses/by/4.0/
The gist: We introduce Equilibrium Matching (EqM), a generative modeling framework built from an equilibrium dynamics perspective.
Terminology
Abstract
We introduce Equilibrium Matching (EqM), a generative modeling framework built from an equilibrium dynamics perspective. EqM discards the non-equilibrium, time-conditional dynamics in traditional diffusion and flow-based generative models and instead learns the equilibrium gradient of an implicit energy landscape. Through this approach, we can adopt an optimization-based sampling process at inference time, where samples are obtained by gradient descent on the learned landscape with adjustable step sizes, adaptive optimizers, and adaptive compute. Empirically, EqM achieves the best FID among the tested methods, reaching 1.90 on ImageNet 256 times 256, while remaining competitive on other metrics. Theoretically, we show under standard assumptions that real images are attracting critical points of the learned field and the energy interpretation applies when the field is conservative. Beyond generation, EqM is a flexible framework that naturally handles tasks including partially noised image denoising, OOD detection, and image composition. By replacing time-conditional velocities with a unified equilibrium landscape, EqM offers a tighter bridge between flow and energy-based models and a simple route to optimization-driven inference.
Sources
- Energy Matching: Unifying Flow Matching and Energy-Based Models for Generative Modeling
- Improved Contrastive Divergence Training of Energy Based Models
- Learning Energy-Based Models by Diffusion Recovery Likelihood
- Your Classifier is Secretly an Energy Based Model and You Should Treat it Like One
- Flow Matching Guide and Code
- PixelCNN++: Improving the PixelCNN with Discretized Logistic Mixture Likelihood and Other Modifications
- Score-Based Generative Modeling through Stochastic Differential Equations
- Is Noise Conditioning Necessary for Denoising Generative Models?
- Diffuse and Disperse: Image Generation with Representation Regularization
Related papers
- A Physics-Regulated Neural Framework for Learning 3D Grain Growth Dynamics
- FlexMoE: Learning One-for-All Multi-Budget Expert Pruning for Flexible MoE Deployment
- OptiMind: Improving Small Language Models for Optimization Modeling through Formulation Guidance and Multi-Turn Learning
- Staying in the Sweet Spot: Boost Reasoning Evolution via Responsive Rollout Difficulty Manipulation
- Examining the Impact of Data and Training Methods on Emergent and Subliminal Misalignment
- Rethinking Black-Box On-Policy Distillation: A Rubric-based Analysis