REPA-G: Test-Time Conditioning with Representation-Aligned Visual Features

Nicolas Sereyjol-Garros   ·   Ellington Kirby   ·   Victor Letzelter   ·   Victor Besnier   ·   Nermin Samet

NeurIPS 2026

Abstract

While representation alignment with self-supervised models has been shown to improve diffusion model training, its potential for enhancing inference-time conditioning remains largely unexplored. We introduce Representation-Aligned Guidance (REPA-G), a framework that leverages these aligned representations, with rich semantic properties, to enable test-time conditioning from features in generation. By optimizing a similarity objective (the potential) at inference, we steer the denoising process toward a conditioned representation extracted from a pre-trained feature extractor. Our method provides versatile control at multiple scales, ranging from fine-grained texture matching via single patches to broad semantic guidance using global image feature tokens. We further extend this to multi-concept composition, allowing for the faithful combination of distinct concepts. REPA-G operates entirely at inference time, offering a flexible and precise alternative to often ambiguous text prompts or coarse class labels. We theoretically justify how this guidance enables sampling from the potential-induced tilted distribution. Quantitative results on ImageNet and COCO demonstrate that our approach achieves high-quality, diverse generations.

BibTeX

@inproceedings{sereyjol2026repag,
  title     = {REPA-G: Test-Time Conditioning with Representation-Aligned Visual Features},
  author    = {Sereyjol-Garros, Nicolas and Kirby, Ellington and Letzelter, Victor and Besnier, Victor and Samet, Nermin},
  booktitle = {Advances in Neural Information Processing Systems (NeurIPS)},
  year      = {2026}
}