Predict to Skip: Linear Multistep Feature Forecasting for Efficient Diffusion Transformers

TL;DR

PrediT introduces a training-free, linear multistep feature prediction framework for diffusion transformers, significantly reducing computation time while maintaining high-quality image and video generation.

Contribution

The paper presents a novel linear multistep prediction method for diffusion transformers that accelerates generation without retraining or quality loss.

Findings

Achieves up to 5.54x latency reduction

Maintains high fidelity in generated images and videos

Effective in both image and video diffusion models

Abstract

Diffusion Transformers (DiT) have emerged as a widely adopted backbone for high-fidelity image and video generation, yet their iterative denoising process incurs high computational costs. Existing training-free acceleration methods rely on feature caching and reuse under the assumption of temporal stability. However, reusing features for multiple steps may lead to latent drift and visual degradation. We observe that model outputs evolve smoothly along much of the diffusion trajectory, enabling principled predictions rather than naive reuse. Based on this insight, we propose \textbf{PrediT}, a training-free acceleration framework that formulates feature prediction as a linear multistep problem. We employ classical linear multistep methods to forecast future model outputs from historical information, combined with a corrector that activates in high-dynamics regions to prevent error accumulation. A dynamic step modulation mechanism adaptively adjusts the prediction horizon by monitoring the feature change rate. Together, these components enable substantial acceleration while preserving generation fidelity. Extensive experiments validate that our method achieves up to $5.54\times$ latency reduction across various DiT-based image and video generation models, while incurring negligible quality degradation.