Scaling Beyond Masked Diffusion Language Models
-
Updated
Feb 18, 2026 - Python
#
diffusion-llms
Here are 2 public repositories matching this topic...
This work introduces Flow Matching Mixture of Experts (FM-MoE), a framework that replaces conventional MLP experts with flow matching networks. Each expert learns a continuous transformation through an ordinary differential equation (ODE), enabling more expressive feature mappings while maintaining the computational efficiency of sparse experts
ai tensorflow transformers pytorch moe tensors diffusion-models flow-matching agents-sdk llm-space diffusion-llms
-
Updated
Feb 4, 2026 - Python
Improve this page
Add a description, image, and links to the diffusion-llms topic page so that developers can more easily learn about it.
Add this topic to your repo
To associate your repository with the diffusion-llms topic, visit your repo's landing page and select "manage topics."