# SpectralDiT: Timestep-Conditioned Spectral Residual Correction for Flow-Matching DiTs Source: [SpectralDiT: Timestep-Conditioned Spectral Residual Correction for Flow-Matching DiTs]() Collection index: [llms.txt]() Data updated: 2026-06-19T05:00:05.274Z Authors: Jiayu Tian Published: 2026-06-17T07:25:24.000Z ## Abstract We propose SpectralDiT, a lightweight modification to flow-matching Diffusion Transformers that adds timestep-conditioned spectral correction to the MLP residual branch. The module decomposes each residual update into low- and high-frequency components on the patch-token grid, then learns a zero-initialized additive gate so the model initially matches the baseline DiT. On CIFAR-10 pixel-space generation, SpectralDiT improves FID from 20.78 to 19.71 at patch size 1 and reduces the radial Fourier spectrum gap. Furthermore, we scale our method to latent diffusion on ImageNet-100. With 0.6% additional theoretical FLOPs and 1.36% additional parameters, SpectralDiT improves latent flow-matching, achieving an 8.7% relative FID reduction under classifier-free guidance (CFG 2.0). All reported results are averaged over five seeds. Ablations and gate visualizations on CIFAR-10 reveal stable block-specific spectral correction patterns. ## Recorded benchmark results A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. - [Image Generation on ImageNet-100 256x256 (50k reference set)](): Image Generation; ImageNet-100 256 x 256 (50,000-sample reference set); FID-50K: 13.59. ## Original sources and implementation [Original paper]() [PDF]()