Skip to content
NewA calmer, faster ZeroHour

Embedding Prediction Helps Image Generation

arXiv cs.AI / cs.LG / cs.CLby Sihan Xu ()1 min read3 · 1 read
AI summary · grok-4.7Summaries by Vani AI

NEPA-DiT-XL predicts embeddings each denoising step and reaches ImageNet FID 1.32 with about one-third of REPA's compute.

Next-Embedding Predictive Autoregression trains a Transformer to predict the next continuous embedding rather than reuse one fixed class or text condition. In Embedding Conditioned Generation, a diffusion transformer is conditioned on Multi-Embedding Prediction outputs that are recomputed at every denoising step. On class-conditional ImageNet 256x256, NEPA-DiT-XL combined with REPA reaches an FID of 1.32 while using about one-third of REPA's training compute, at the cost of a second network on each sampling step.

Key facts
  • NEPA predicts next continuous embeddings instead of reusing a fixed class or text condition.
  • Multi-Embedding Prediction forecasts clean-image embeddings from the condition and noisy image.
  • The diffusion generator is reconditioned on those predictions at every denoising step.
  • NEPA-DiT-XL reaches FID 1.32 on ImageNet 256 using about one-third of REPA compute.
Explain this AI news with Vani AIWhat was announced, what is new and why it matters, in a new chat

From the original article

1 min read at arxiv.org

In diffusion transformers, a class label or a text prompt is embedded once, and the same condition is reused at every denoising step. We ask whether predicted embeddings can serve as this condition instead. Next-Embedding Predictive Autoregression (NEPA) trains a Transformer to predict the next continuous embedding in a sequence.

arXiv cs.AI / cs.LG / cs.CL

By Sihan Xu · arxiv.org

Read the full article at arXiv cs.AI / cs.LG / cs.CL

A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.