DIFF.BLOG
New Following Discover Jobs
More
Top Writers Suggest a blog Upvotes plugin
Report bug Contact About Privacy
Sign up
Topics
Follow your own topics →
Menu
New Following Discover Jobs Top Writers
More
Suggest a blog Upvotes plugin Report bug Contact About
Sign up
Privacy Policy
The home for great developer writing.
We surface the best developer writing from thousands of independent blogs, updated daily.
Join Diff.blog
TOPICS

Test-time scaling of diffusions with flow maps

· Research Nvidia · Nov. 1, 2026, 3:16 a.m.
Machine Learning Diffusion models Image Editing Algorithm Improvements
Summary
The post introduces a new method called Flow Map Trajectory Tilting (FMTT), which improves diffusion models during test-time by directly utilizing flow maps to incorporate user-specified rewards into the diffusion dynamics, leading to better sampling and local maximization of reward distributions.
Read full post on research.nvidia.com →
MORE POSTS LIKE THIS
DiLaDiff: Distilled Latent-Augmented Diffusion for Language Modeling
Research Nvidia · Nov 1, 2026
Machine Learning NVIDIA
AI Project: How to Edit Images with AI (Inpainting with diffusers)
Ahmed Nabil · May 6, 2026
Data Science python projects
An Introduction to Image Generation with Diffusion Models (2/2)
Mikel Sagardia · Jan 22, 2026
AI engineering
Optimizing Transformer-Based Diffusion Models for Video Generation with NVIDIA TensorRT
NVIDIA Corporation · Apr 21, 2025
AI Platforms / Deployment Development & Optimization
What inference-time scaling should be like for Diffusion Model
Norm Inui · Jan 18, 2025
Sparks Machine Learning
Diffusion Models Background and Code
Mahesh Bhosale · Nov 28, 2024
Technical diffusion
Discover more posts →
AUTHOR
RECENT POSTS FROM THE AUTHOR
Choose how you want to continue.
Continue with GitHub Continue with Google
By continuing, you agree to our Privacy Policy.