# Deep Residual Mixture Models

arXiv: https://arxiv.org/abs/2006.12063 · HTML: https://mlsanity.com/paper/2006.12063/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 22 Jun 2020.

## Built on by (most cited first, 4 of 4)

- [MAUVE: Measuring the Gap Between Neural Text and Human Text using Divergence Frontiers](https://mlsanity.com/paper/2102.01454/) (2021)
- [MAUVE Scores for Generative Models: Theory and Practice](https://mlsanity.com/paper/2212.14578/) (2022)
- [Divergence Frontiers for Generative Models: Sample Complexity, Quantization Effects, and Frontier Integrals](https://mlsanity.com/paper/2106.07898/) (2021)
- [Deep Skew-t Mixture Models](https://mlsanity.com/paper/2609.00773/) (2026)

## Abstract (arXiv)

We propose Deep Residual Mixture Models (DRMMs), a novel deep generative model architecture. Compared to other deep models, DRMMs allow more flexible conditional sampling: The model can be trained once with all variables, and then used for sampling with arbitrary combinations of conditioning variables, Gaussian priors, and (in)equality constraints. This provides new opportunities for interactive and exploratory machine learning, where one should minimize the user waiting for retraining a model. We demonstrate DRMMs in constrained multi-limb inverse kinematics and controllable generation of animations.