BestHub
Sign in
Tagged articles

one-step generation

4 articles · Page 1 of 1
Machine Learning Algorithms & Natural Language Processing
Machine Learning Algorithms & Natural Language Processing
Aug 10, 2026 · Artificial Intelligence

Training One‑Step Generative Models Without CFG, DMD, GAN, Drifting, or MeanFlow

The paper introduces TBSM, a lightweight direction‑tracking network that learns fake‑to‑real movement fields from three‑sample scattering events, enabling single‑forward‑pass generation and achieving state‑of‑the‑art FID scores on ImageNet and a 20B text‑to‑image model without relying on CFG, DMD, GAN, drifting, or mean‑flow techniques.

PixelDiTQwen-ImageTBSM
0 likes · 12 min read
Training One‑Step Generative Models Without CFG, DMD, GAN, Drifting, or MeanFlow
Machine Heart
Machine Heart
Aug 9, 2026 · Artificial Intelligence

How to Train a One‑Step Generative Model Without CFG, DMD, GAN, or Drifting

The paper introduces TBSM, a lightweight direction‑tracking network that learns per‑sample Fake‑to‑Real vectors to guide a generator, enabling single‑forward (NFE=1) image synthesis with FID 1.92 on ImageNet‑512 and high‑quality text‑to‑image results, all without CFG, DMD, GAN, or drifting methods.

TBSMdiffusion modelsdirection tracking
0 likes · 11 min read
How to Train a One‑Step Generative Model Without CFG, DMD, GAN, or Drifting
Machine Learning Algorithms & Natural Language Processing
Machine Learning Algorithms & Natural Language Processing
Jul 8, 2026 · Artificial Intelligence

Why the One‑Step “Drift Model” by He Kaiming’s Team Was Rejected at ICML (Scores 5‑4‑4‑3)

The drift model moves the distribution evolution into training to achieve single‑step image synthesis with near‑SOTA FID scores, but ICML reviewers gave it 5/4/4/3 and rejected it, citing heavy reliance on a pretrained feature encoder, unclear contribution of the drift field versus classifier‑free guidance, and insufficient comparison to prior work.

FIDICML rejectionclassifier-free guidance
0 likes · 9 min read
Why the One‑Step “Drift Model” by He Kaiming’s Team Was Rejected at ICML (Scores 5‑4‑4‑3)
AIWalker
AIWalker
Mar 4, 2026 · Artificial Intelligence

Drifting Models Enable One‑Step Generation, Shattering Speed Records

The paper introduces Drifting Models, a new generative paradigm that moves the distribution evolution to the training phase, achieving true one‑step (1‑NFE) generation with state‑of‑the‑art ImageNet FID scores of 1.54 in latent space and 1.61 in pixel space, while eliminating the need for distillation or classifier‑free guidance.

Drifting ModelsImageNetdiffusion
0 likes · 24 min read
Drifting Models Enable One‑Step Generation, Shattering Speed Records