Optimal Mixture-of-Experts Model Averaging for Conditional Generative Models
Shijin Gong, Baihua He, Xinyu Zhang
stat.ML, cs.LG
Submitted: 2026-07-05
Comments: 32 pages, 1 figures
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Terminology
Sources
- eDiff-I: Text-to-Image Diffusion Models with an Ensemble of Expert Denoisers
- A Survey on Mixture of Experts in Large Language Models
- Deep conditional distribution learning via conditional F"ollmer flow
- Combining pre-trained models via localized model averaging
- Prediction-powered Inference by Mixture of Experts
- Diffusion Boosted Trees
- CARD: Classification and Regression Diffusion Models
- Evaluating Aleatoric Uncertainty via Conditional Generative Models
- TabDDPM: Modelling Tabular Data with Diffusion Models
- TabResFlow: A Normalizing Spline Flow Model for Probabilistic Univariate Tabular Regression
- Conditional Generative Adversarial Nets
- GPT-4 Technical Report
- Outrageously Large Neural Networks: The Sparsely-Gated Mixture-of-Experts Layer
Related papers
- Behavior of prediction performance metrics with rare events
- Optimal Estimation of Generic Dynamics by Path-Dependent Neural Jump ODEs
- A Posterior-Dynamics Framework for Imaging Inverse Problems with Pretrained Diffusion Priors
- Scalable Posterior Uncertainty for Flexible Density-Based Clustering
- A Transport Geometry of Benign Overfitting
- Efficient quadratic entropy with distance sketches