Year: 2026
Type: preprint
Abstract: We aim to improve frozen flow-matching image generators by adding inference computation inside the denoiser, without changing model weights or the outer sampler. Existing generators usually spend extra test-time computation by increasing the number of sampling steps, which repeatedly evaluates the entire denoiser and couples quality gains to sampler cost. A key challenge is how to use extra computation inside a frozen transformer denoiser: the method must decide which tokens, layers, and sampling times receive repeated updates while preserving the original generation pipeline. We introduce a training-free looping framework that repeatedly applies selected transformer layers inside each denoising call. Dense and Sparse Token Loop vary the token scope; Sampling-Progress Gating and the loop layer range specify when and where looping is active; loop count and strength control the repeated updates; and Loop Guidance combines ordinary and looped vector-field predictions. Across two Scale-RAE model scales, loop variants improve primary and auxiliary quality metrics with competitive quality--efficiency trade-offs. Loop Guidance further improves both primary metrics across all three tested models; on Scale-RAE DiT2.4B, it raises GenEval from 0.4471 to 0.5691 and DPG-Bench from 0.7656 to 0.8053. Code will be released.
Source: arXiv (Cornell University)
Authors: Yan, Yuanyi, Rao, Xinzhe, Shen, Canyu, Chen, Yang, Chen, Yunlu, Tang, Meng, Long, Teng, Hu, Vincent Tao
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7204964823
Cites:
Cited by:
Related to: