# Exponential Tilting of Generative Models: Improving Sample Quality by Training and Sampling from Latent Energy

arXiv: https://arxiv.org/abs/2006.08100 · HTML: https://mlsanity.com/paper/2006.08100/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 15 Jun 2020.

## Built on by (most cited first, 6 of 6)

- [Contrastive Energy Prediction for Exact Energy-Guided Diffusion Sampling in Offline Reinforcement Learning](https://mlsanity.com/paper/2304.12824/) (2023)
- [Bi-level Doubly Variational Learning for Energy-based Latent Variable Models](https://mlsanity.com/paper/2203.14702/) (2022)
- [Energy-Based Models for Functional Data using Path Measure Tilting](https://mlsanity.com/paper/2202.01929/) (2022)
- [Balanced Training of Energy-Based Models with Adaptive Flow Sampling](https://mlsanity.com/paper/2306.00684/) (2023)
- [EBMs Trained with Maximum Likelihood are Generator Models Trained with a Self-adverserial Loss](https://mlsanity.com/paper/2102.11757/) (2021)
- [One Diffusion Model, Two Roles: Guided Trajectory Planning and Safety-Critical Scenario Generation in Closed-Loop Simulation](https://mlsanity.com/paper/2609.04921/) (2026)

## Abstract (arXiv)

In this paper, we present a general method that can improve the sample quality of pre-trained likelihood based generative models. Our method constructs an energy function on the latent variable space that yields an energy function on samples produced by the pre-trained generative model. The energy based model is efficiently trained by maximizing the data likelihood, and after training, new samples in the latent space are generated from the energy based model and passed through the generator to producing samples in observation space. We show that using our proposed method, we can greatly improve the sample quality of popular likelihood based generative models, such as normalizing flows and VAEs, with very little computational overhead.