Lune

CVPR Findings2026

FREE: Uncertainty-Aware Autoregression for Parallel Diffusion Transformers

Xinwan Wen, Bowen Li, Jiajun Luo, Ye Li, Zhi Wang

Abstract

To address this limitation, we analyze the DiTs' feature dynamics and find the features of the final transformer layer (top-block) exhibit strong temporal consistency and rich semantic abstraction. Based on this insight, we propose FREE, a novel framework that employs a lightweight drafter to perform featurelevel autoregression with parallel verification, guaranteeing lossless acceleration with theoretical and empirical support. Meanwhile, prediction variance (uncertainty) of DiTs naturally increases in later denoising steps, reducing acceptance rates under speculative sampling. To mitigate this effect, we further introduce an uncertainty-guided relaxation strategy, forming FREE (relax), which dynamically adjusts the acceptance probability in response to uncertainty levels. Experiments on ImageNet-512 2 show that FREE achieves up to 1.86× acceleration, and FREE (relax) further reaches 2.25× speedup while maintaining high perceptual and quantitative fidelity in generation quality.

Ask Lune

Know this paper in minutes, not an afternoon.

Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.

Questions to start from

Lune can draw on

the full paper · related work

Get the answer

Free to start. No credit card required.

Lineage

  • High-Resolution Image Synthesis with Latent Diffusion Models cites Denoising Diffusion Probabilistic Models
  • High-Resolution Image Synthesis with Latent Diffusion Models cites Denoising Diffusion Implicit Models
  • Scalable Diffusion Models with Transformers cites High-Resolution Image Synthesis with Latent Diffusion Models
  • Scalable Diffusion Models with Transformers cites Denoising Diffusion Probabilistic Models
  • Scalable Diffusion Models with Transformers cites Denoising Diffusion Implicit Models
  • Denoising Diffusion Implicit Models cites Denoising Diffusion Probabilistic Models
Builds onCited byLeft to right by year; the more cited, the higher.

Builds on37

  1. Denoising Diffusion Probabilistic ModelsJonathan Ho, Ajay Jain, Pieter AbbeelNeurIPS 2020 · 35,902 citations
  2. High-Resolution Image Synthesis with Latent Diffusion ModelsRobin Rombach, Andreas Blattmann, Dominik Lorenz, Patrick Esser et al.CVPR 2022 · 13,123 citations
  3. Denoising Diffusion Implicit ModelsJiaming Song, Chenlin Meng, Stefano ErmonICLR 2021 · 11,743 citations
  4. Scalable Diffusion Models with TransformersWilliam Peebles, Saining XieICCV 2023 · 5,568 citations
  5. FlashAttention: Fast and Memory-Efficient Exact Attention with IO-AwarenessTri Dao, Daniel Y. Fu, Stefano Ermon, Atri Rudra et al.NeurIPS 2022 · 5,493 citations

Related papers

  1. DFlash: Block Diffusion for Flash Speculative DecodingJian Chen, Yesheng Liang, Zhijian LiuICML 2026
  2. VVS: Accelerating Speculative Decoding for Visual Autoregressive Generation via Partial Verification SkippingHaotian Dong, Ye Li, Rongwei Lu, Chen Tang et al.CVPR 2026 · 4 citations
  3. Speculative Jacobi-Denoising Decoding for Accelerating Autoregressive Text-to-image GenerationYao Teng, Fuyun Wang, Xian Liu, Zhekai Chen et al.NeurIPS 2025 · 8 citations
  4. SpecBound: Adaptive Bounded Self-Speculation with Layer-wise Confidence CalibrationZhuofan Wen, Yang FengACL Findings 2026
  5. Sortblock: Similarity-Aware Feature Reuse for Diffusion ModelHanqi Chen, Xu Zhang, Xiaoliu Guan, Lielin Jiang et al.AAAI 2026
Dusk over the sea between two cliffs drawn in fine vertical lines