Diffusion Model Inference Acceleration

Follow this topic Ask about this topic

Momentum

40 papers in the last four weeks, up 100% on the four weeks before. 0.4% of all new papers.

Jul 13Week of Sep 28

Latest papers 335

Apr 20, 2026cs.CV

AdaCluster: Adaptive Query-Key Clustering for Sparse Attention in Video Generation

Video diffusion transformers (DiTs) suffer from prohibitive inference latency due to quadratic attention complexity. Existing sparse attention methods either overlook semantic similarity or fail to adapt to heterogeneous token distributions across layers, leading to model performance degradation. We propose AdaCluster, a training-free adaptive clustering framework that accelerates the generation of DiTs while preserving accuracy. AdaCluster applies an angle-similarity-preserving clustering method to query vectors for higher compression, and designs a euclidean-similarity-preserving clustering method for keys, covering cluster number assignment, threshold-wise adaptive clustering, and efficient critical cluster selection. Experiments on CogVideoX-2B, HunyuanVideo, and Wan-2.1 on one A40 GPU demonstrate up to 1.67-4.31x speedup with negligible quality degradation.

Haoyue Tan, Shengnan Wang, Yulin Qiao +5

University of Science and Technology of China · Institute of Artificial Intelligence, Hefei Comprehensive National Science Center · 3Independent Researcher +2

Video Diffusion Models Diffusion Transformer Sparse Attention Diffusion Model Inference Acceleration

Apr 20, 2026cs.LG

Efficient Diffusion Models under Nonconvex Equality and Inequality constraints via Landing

Generative modeling within constrained sets is essential for scientific and engineering applications involving physical, geometric, or safety requirements (e.g., molecular generation, robotics). We present a unified framework for constrained diffusion models on generic nonconvex feasible sets

Σ

that simultaneously enforces equality and inequality constraints throughout the diffusion process. Our framework incorporates both overdamped and underdamped dynamics for forward and backward sampling. A key algorithmic innovation is a computationally efficient landing mechanism that replaces costly and often ill-defined projections onto

Σ

, ensuring feasibility without iterative Newton solves or projection failures. By leveraging underdamped dynamics, we accelerate mixing toward the prior distribution, effectively alleviating the high simulation costs typically associated with constrained diffusion. Empirically, this approach reduces function evaluations and memory usage during both training and inference while preserving sample quality. On benchmarks featuring equality and mixed constraints, our method achieves comparable sample quality to state-of-the-art baselines while significantly reducing computational cost, providing a practical and scalable solution for diffusion on nonconvex feasible sets.

Kijung Jeon, Michael Muehlebach, Molei Tao

Georgia Institute of Technology · Max Planck Institute for Int. Sys.

Constrained Generative Modeling Langevin Dynamics Diffusion Model Sampling Diffusion Model Inference Acceleration Diffusion Models

Apr 19, 2026cs.CV

Speculative Decoding for Autoregressive Video Generation

Autoregressive video diffusion is emerging as a promising paradigm for streaming video synthesis, with step distillation serving as the primary means of accelerating inference. Whether speculative decoding, the dominant acceleration strategy for large language models, can be effectively adapted to autoregressive video generation remains an open question, because video blocks are continuous spatiotemporal tensors with no token-level distribution for exact rejection sampling. We introduce SDVG, which brings speculative decoding to block-based autoregressive video diffusion by replacing token verification with an image-quality router. A 1.3B drafter proposes candidate blocks via four denoising steps; each block is VAE-decoded and scored by ImageReward using worst-frame aggregation--taking the minimum per-frame reward to catch single-frame artifacts that averaging would mask. Blocks scoring above a fixed threshold tau are accepted into the 14B target's KV cache; the rest are regenerated by the target. Two additional design choices prove critical: the first block is always force-rejected to anchor scene composition, and tau serves as a single knob that traces a smooth quality-speed Pareto frontier. On 1003 MovieGenVideoBench prompts (832x480), SDVG retains 98.1% of target-only VisionReward quality (0.0773 vs. 0.0788) at a 1.59x speedup with tau=-0.7, and reaches 2.09x at 95.7% quality retention--while consistently outperforming draft-only generation by over +17%. The framework is training-free, requires no architectural changes, and can be seamlessly integrated into existing autoregressive video generation pipelines.

Yuezhou Hu, Jintao Zhang

University of California, Berkeley

Video Diffusion Models Autoregressive Diffusion Speculative Decoding Autoregressive Video Generation Diffusion Model Inference Acceleration

Apr 17, 2026cs.RO

VADF: Vision-Adaptive Diffusion Policy Framework for Efficient Robotic Manipulation

Diffusion policies are becoming mainstream in robotic manipulation but suffer from hard negative class imbalance due to uniform sampling and lack of sample difficulty awareness, leading to slow training convergence and frequent inference timeout failures. We propose VADF (Vision-Adaptive Diffusion Policy Framework), a vision-driven dual-adaptive framework that significantly reduces convergence steps and achieves early success in inference, with model-agnostic design enabling seamless integration into any diffusion policy architecture. During training, we introduce Adaptive Loss Network (ALN), a lightweight MLP-based loss predictor that quantifies per-step sample difficulty in real time. Guided by hard negative mining, it performs weighted sampling to prioritize high-loss regions, enabling adaptive weight updates and faster convergence. In inference, we design the Hierarchical Vision Task Segmenter (HVTS), which decomposes high-level task instructions into multi-stage low-level sub-instructions based on visual input. It adaptively segments action sequences into simple and complex subtasks by assigning shorter noise schedules with longer direct execution sequences to simple actions, and longer noise steps with shorter execution sequences to complex ones, thereby dramatically reducing computational overhead and significantly improving the early success rate.

Xinglei Yu, Zhenyang Liu, Shufeng Nan +2

Fudan University · Shanghai Innovation Institute

Adaptive Inference Diffusion Policy Robotic Manipulation Diffusion Model Inference Acceleration

Apr 17, 2026cs.CV

Efficient Video Diffusion Models: Advancements and Challenges

Video diffusion models have rapidly become the dominant paradigm for high-fidelity generative video synthesis, but their practical deployment remains constrained by severe inference costs. Compared with image generation, video synthesis compounds computation across spatial-temporal token growth and iterative denoising, making attention and memory traffic major bottlenecks in real-world settings. This survey provides a systematic and deployment-oriented review of efficient video diffusion models. We propose a unified categorization that organizes existing methods into four classes of main paradigms, including step distillation, efficient attention, model compression, and cache/trajectory optimization. Building on this categorization, we respectively analyze algorithmic trends of these four paradigms and examine how different design choices target two core objectives: reducing the number of function evaluations and minimizing per-step overhead. Finally, we discuss open challenges and future directions, including quality preservation under composite acceleration, hardware-software co-design, robust real-time long-horizon generation, and open infrastructure for standardized evaluation. To the best of our knowledge, our work is the first comprehensive survey on efficient video diffusion models, offering researchers and engineers a structured overview of the field and its emerging research directions.

Shitong Shao, Lichen Bai, Pengfei Wan +2

The Hong Kong University of Science and Technology (Guangzhou), China · Kling Team, Kuaishou Technology, China · The Hong Kong University of Science and Technology, Hong Kong

Model Compression Video Diffusion Models Diffusion Model Distillation Efficient Attention Diffusion Model Caching Diffusion Model Inference Acceleration

Mar 30, 2026cs.CV

Drift-AR: Single-Step Visual Autoregressive Generation via Anti-Symmetric Drifting

Autoregressive (AR)-Diffusion hybrid paradigms combine AR's structured semantic modeling with diffusion's high-fidelity synthesis, yet suffer from a dual speed bottleneck: the sequential AR stage and the iterative multi-step denoising of the diffusion vision decode stage. Existing methods address each in isolation without a unified principle design. We observe that the per-position \emph{prediction entropy} of continuous-space AR models naturally encodes spatially varying generation uncertainty, which simultaneously governing draft prediction quality in the AR stage and reflecting the corrective effort required by vision decoding stage, which is not fully explored before. Since entropy is inherently tied to both bottlenecks, it serves as a natural unifying signal for joint acceleration. In this work, we propose \textbf{Drift-AR}, which leverages entropy signal to accelerate both stages: 1) for AR acceleration, we introduce Entropy-Informed Speculative Decoding that align draft-target entropy distributions via a causal-normalized entropy loss, resolving the entropy mismatch that causes excessive draft rejection; 2) for visual decoder acceleration, we reinterpret entropy as the \emph{physical variance} of the initial state for an anti-symmetric drifting field -- high-entropy positions activate stronger drift toward the data manifold while low-entropy positions yield vanishing drift -- enabling single-step (1-NFE) decoding without iterative denoising or distillation. Moreover, both stages share the same entropy signal, which is computed once with no extra cost. Experiments on MAR, TransDiff, and NextStep-1 demonstrate 3.8-5.5

\times

speedup with genuine 1-NFE decoding, matching or surpassing original quality. Code will be available at https://github.com/aSleepyTree/Drift-AR.

Zhen Zou, Xiaoxiao Ma, Mingde Yao +3

University of Science and Technology of China · The Chinese University of Hong Kong, MMLAB · JD Explore Academy +1

Autoregressive Diffusion Visual Autoregressive Models One-Step Generative Modeling Speculative Decoding Diffusion Model Inference Acceleration

Mar 15, 2026cs.CV

LatSearch: Latent Reward-Guided Search for Faster Inference-Time Scaling in Video Diffusion

The recent success of inference-time scaling in large language models has inspired similar explorations in video diffusion. In particular, motivated by the existence of "golden noise" that enhances video quality, prior work has attempted to improve inference by optimising or searching for better initial noise. However, these approaches have notable limitations: they either rely on priors imposed at the beginning of noise sampling or on rewards evaluated only on the denoised and decoded videos. This leads to error accumulation, delayed and sparse reward signals, and prohibitive computational cost, which prevents the use of stronger search algorithms. Crucially, stronger search algorithms are precisely what could unlock substantial gains in controllability, sample efficiency and generation quality for video diffusion, provided their computational cost can be reduced. To fill in this gap, we enable efficient inference-time scaling for video diffusion through latent reward guidance, which provides intermediate, informative and efficient feedback along the denoising trajectory. We introduce a latent reward model that scores partially denoised latents at arbitrary timesteps with respect to visual quality, motion quality, and text alignment. Building on this model, we propose LatSearch, a novel inference-time search mechanism that performs Reward-Guided Resampling and Pruning (RGRP). In the resampling stage, candidates are sampled according to reward-normalised probabilities to reduce over-reliance on the reward model. In the pruning stage, applied at the final scheduled step, only the candidate with the highest cumulative reward is retained, improving both quality and efficiency. We evaluate LatSearch on the VBench-2.0 benchmark and demonstrate that it consistently improves video generation across multiple evaluation dimensions compared to the baseline Wan2.1 model.

Zengqun Zhao, Ziquan Liu, Yu Cao +5

Queen Mary University of London · Huawei R&D UK · Imperial College London

Video Diffusion Models Inference-Time Scaling Diffusion Model Inference Acceleration

Mar 13, 2026cs.CV

AccelAes: Accelerating Diffusion Transformers for Training-Free Aesthetic-Enhanced Image Generation

Diffusion Transformers (DiTs) are a dominant backbone for high-fidelity text-to-image generation due to strong scalability and alignment at high resolutions. However, quadratic self-attention over dense spatial tokens leads to high inference latency and limits deployment. We observe that denoising is spatially non-uniform with respect to aesthetic descriptors in the prompt. Regions associated with aesthetic tokens receive concentrated cross-attention and show larger temporal variation, while low-affinity regions evolve smoothly with redundant computation. Based on this insight, we propose AccelAes, a training-free framework that accelerates DiTs through aesthetics-aware spatio-temporal reduction while improving perceptual aesthetics. AccelAes builds AesMask, a one-shot aesthetic focus mask derived from prompt semantics and cross-attention signals. When localized computation is feasible, SkipSparse reallocates computation and guidance to masked regions. We further reduce temporal redundancy using a lightweight step-level prediction cache that periodically replaces full Transformer evaluations. Experiments on representative DiT families show consistent acceleration and improved aesthetics-oriented quality. On Lumina-Next, AccelAes achieves a 2.11

\times

speedup and improves ImageReward by +11.9% over the dense baseline. Code is available at https://github.com/xuanhuayin/AccelAes.

Xuanhua Yin, Chuanzhi Xu, Haoxian Zhou +2

School of Computer Science, The University of Sydney, NSW 2006, Australia

Diffusion Transformer Image Generation Attention Control in Diffusion Models Diffusion Model Inference Acceleration T2I Diffusion Models Diffusion Models

Mar 9, 2026cs.CV

SVG-EAR: Parameter-Free Linear Compensation for Sparse Video Generation via Error-aware Routing

Diffusion Transformers (DiTs) have become a leading backbone for video generation, yet their quadratic attention cost remains a major bottleneck. Sparse attention reduces this cost by computing only a subset of attention blocks. However, prior methods often either drop the remaining blocks, which incurs information loss, or rely on learned predictors to approximate them, introducing training overhead and potential output distribution shifting. In this paper, we show that the missing contributions can be recovered without training: after semantic clustering, keys and values within each block exhibit strong similarity and can be well summarized by a small set of cluster centroids. Based on this observation, we introduce SVG-EAR, a parameter-free linear compensation branch that uses the centroid to approximate skipped blocks and recover their contributions. While centroid compensation is accurate for most blocks, it can fail on a small subset. Standard sparsification typically selects blocks by attention scores, which indicate where the model places its attention mass, but not where the approximation error would be largest. SVG-EAR therefore performs error-aware routing: a lightweight probe estimates the compensation error for each block, and we compute exactly the blocks with the highest error-to-cost ratio while compensating for skipped blocks. We provide theoretical guarantees that relate attention reconstruction error to clustering quality, and empirically show that SVG-EAR improves the quality-efficiency trade-off and increases throughput at the same generation fidelity on video diffusion tasks. Overall, SVG-EAR establishes a clear Pareto frontier over prior approaches, achieving up to 1.77

\times

and 1.93

\times

speedups while maintaining PSNRs of up to 29.759 and 31.043 on Wan2.2 and HunyuanVideo, respectively.

Xuanyi Zhou, Qiuyang Mang, Shuo Yang +7

UC Berkeley, USA

Video Diffusion Models Efficient Attention Sparse Attention Diffusion Model Inference Acceleration Block-Sparse Attention

Mar 6, 2026cs.CV

Cross-Resolution Distribution Matching for Diffusion Distillation

Diffusion distillation is central to accelerating image and video generation, yet existing methods are fundamentally limited by the denoising process, where step reduction has largely saturated. Partial timestep low-resolution generation can further accelerate inference, but it suffers noticeable quality degradation due to cross-resolution distribution gaps. We propose Cross-Resolution Distribution Matching Distillation (RMD), a novel distillation framework that bridges cross-resolution distribution gaps for high-fidelity, few-step multi-resolution cascaded inference. Specifically, RMD divides the timestep intervals for each resolution using logarithmic signal-to-noise ratio (logSNR) curves, and introduces logSNR-based mapping to compensate for resolution-induced shifts. Distribution matching is conducted along resolution trajectories to reduce the gap between low-resolution generator distributions and the teacher's high-resolution distribution. In addition, a predicted-noise re-injection mechanism is incorporated during upsampling to stabilize training and improve synthesis quality. Quantitative and qualitative results show that RMD preserves high-fidelity generation while accelerating inference across various backbones. Notably, RMD achieves up to 33.4X speedup on SDXL and 25.6X on Wan2.1-14B, while preserving high visual fidelity.

Feiyang Chen, Hongpeng Pan, Haonan Xu +3

Huawei, Shanghai & Hangzhou, China · Nanjing University of Science and Technology, Nanjing, China

Video Diffusion Models Diffusion Model Distillation Few-Step Diffusion Sampling Diffusion-Based Image Generation Diffusion Model Inference Acceleration

Mar 2, 2026cs.CV

Adaptive Spectral Feature Forecasting for Diffusion Sampling Acceleration

Diffusion models have become the dominant tool for high-fidelity image and video generation, yet are critically bottlenecked by their inference speed due to the numerous iterative passes of Diffusion Transformers. To reduce the exhaustive compute, recent works resort to the feature caching and reusing scheme that skips network evaluations at selected diffusion steps by using cached features in previous steps. However, their preliminary design solely relies on local approximation, causing errors to grow rapidly with large skips and leading to degraded sample quality at high speedups. In this work, we propose spectral diffusion feature forecaster (Spectrum), a training-free approach that enables global, long-range feature reuse with tightly controlled error. In particular, we view the latent features of the denoiser as functions over time and approximate them with Chebyshev polynomials. Specifically, we fit the coefficient for each basis via ridge regression, which is then leveraged to forecast features at multiple future diffusion steps. We theoretically reveal that our approach admits more favorable long-horizon behavior and yields an error bound that does not compound with the step size. Extensive experiments on various state-of-the-art image and video diffusion models consistently verify the superiority of our approach. Notably, we achieve up to 4.79

\times

speedup on FLUX.1 and 4.67

\times

speedup on Wan2.1-14B, while maintaining much higher sample quality compared with the baselines.

Jiaqi Han, Juntong Shi, Puheng Li +3

Stanford University · ByteDance

Video Diffusion Models Diffusion-Based Image Generation Diffusion Model Caching Diffusion Model Inference Acceleration

Feb 27, 2026cs.CV

SenCache: Accelerating Diffusion Model Inference via Sensitivity-Aware Caching

Diffusion models achieve state-of-the-art video generation quality, but their inference remains expensive due to the large number of sequential denoising steps. This has motivated a growing line of research on accelerating diffusion inference. Among training-free acceleration methods, caching reduces computation by reusing previously computed model outputs across timesteps. Existing caching methods rely on heuristic criteria to choose cache/reuse timesteps and require extensive tuning. We address this limitation with a principled sensitivity-aware caching framework. Specifically, we formalize the caching error through an analysis of the model output sensitivity to perturbations in the denoising inputs, i.e., the noisy latent and the timestep, and show that this sensitivity is a key predictor of caching error. Based on this analysis, we propose Sensitivity-Aware Caching (SenCache), a dynamic caching policy that adaptively selects caching timesteps on a per-sample basis. Our framework provides a theoretical basis for adaptive caching, explains why prior empirical heuristics can be partially effective, and extends them to a dynamic, sample-specific approach. Experiments on Wan 2.1, CogVideoX, and LTX-Video show that SenCache achieves better visual quality than existing caching methods under similar computational budgets.

Yasaman Haghighi, Alexandre Alahi

Ecole Polytechnique Fédérale de Lausanne (EPFL)

Video Diffusion Models Diffusion Model Caching Diffusion Model Inference Acceleration

Feb 11, 2026cs.LG

General and Efficient Steering of Diffusion Models

Steering diffusion models toward conditions unseen during training typically requires either retraining with conditional inputs or per-step gradient computations, both of which incur substantial computational overhead. We present Noise-Aligned RFM Steering (NA-RFM), a general recipe for efficiently steering diffusion models without gradient guidance during inference, enabling fast controllable generation. The method combines two offline-computed signals: noise alignment, a high-noise correction from PCA statistics of the target examples and the full data, and Recursive Feature Machine (RFM) activation steering, which learns a target-discriminative direction from labeled forward-process activations. During sampling, noise alignment provides coarse control at high noise, while the RFM direction is reused over intermediate/late timesteps through lightweight activation edits. Experiments on CIFAR-10, ImageNet, CelebA, and fine-grained bird species show improved target accuracy over gradient-based post-hoc guidance baselines, improved FID on the class-guidance benchmarks, and substantial inference speedups. Code: https://github.com/isotrivial/na-rfm.

Qingsong Wang, Mikhail Belkin, Yusu Wang

Halıcıoğlu Data Science Institute, University of California San Diego, La Jolla, CA, USA

Diffusion Model Guidance Activation Steering Controllable Image Generation Diffusion Model Inference Acceleration

Feb 5, 2026cs.CV

Stable Velocity: A Variance Perspective on Flow Matching

While flow matching is elegant, its reliance on single-sample conditional velocities leads to high-variance training targets that destabilize optimization and slow convergence. By explicitly characterizing this variance, we identify 1) a high-variance regime near the prior, where optimization is challenging, and 2) a low-variance regime near the data distribution, where conditional and marginal velocities nearly coincide. Leveraging this insight, we propose Stable Velocity, a unified framework that improves both training and sampling. For training, we introduce Stable Velocity Matching (StableVM), an unbiased variance-reduction objective, along with Variance-Aware Representation Alignment (VA-REPA), which adaptively strengthen auxiliary supervision in the low-variance regime. For inference, we show that dynamics in the low-variance regime admit closed-form simplifications, enabling Stable Velocity Sampling (StableVS), a finetuning-free acceleration. Extensive experiments on ImageNet

256\times256

and large pretrained text-to-image and text-to-video models, including SD3.5, Flux, Qwen-Image, and Wan2.2, demonstrate consistent improvements in training efficiency and more than

2\times

faster sampling within the low-variance regime without degrading sample quality. Our code is available at https://github.com/linYDTHU/StableVelocity.

Donglin Yang, Yongxing Zhang, Xin Yu +5

University of Hong Kong, Hong Kong · University of British Columbia, Canada · Vector Institute for AI, Toronto, Canada +2

Flow Matching Image Generation Flow-Based Generative Modeling Variance Reduction Diffusion Model Inference Acceleration

Feb 5, 2026cs.CV

FlashBlock: Attention Caching for Efficient Long-Context Block Diffusion

Generating long-form content, such as minute-long videos and extended texts, is increasingly important for modern generative models. Block diffusion improves inference efficiency via KV caching and block-wise causal inference and has been widely adopted in diffusion language models and video generation. However, in long-context settings, block diffusion still incurs substantial overhead from repeatedly computing attention over a growing KV cache. We identify an underexplored property of block diffusion: cross-step redundancy of attention within a block. Our analysis shows that attention outputs from tokens outside the current block remain largely stable across diffusion steps, while block-internal attention varies significantly. Based on this observation, we propose FlashBlock, a cached block-external attention mechanism that reuses stable attention output, reducing attention computation and KV cache access without modifying the diffusion process. Moreover, FlashBlock is orthogonal to sparse attention and can be combined as a complementary residual reuse strategy, substantially improving model accuracy under aggressive sparsification. Experiments on diffusion language models and video generation demonstrate up to 1.44

\times

higher token throughput and up to 1.6

\times

reduction in attention time, with negligible impact on generation quality. Project page: https://caesarhhh.github.io/FlashBlock/.

Zhuokun Chen, Jianfei Cai, Bohan Zhuang

1Monash University · 2Zhejiang University

Video Diffusion Models Self-Attention Blockwise Diffusion Language Models Efficient Attention Diffusion Model Caching Attention Mechanisms+2

Jan 28, 2026cs.LG

Window-Diffusion: Accelerating Diffusion Language Model Inference with Windowed Token Pruning and Caching

Diffusion language models (DLMs) generate text through iterative denoising, but inference requires full-sequence attention at every iteration, resulting in substantial redundant computation on masked tokens. Block-wise diffusion can reduce this cost, yet it typically relies on retraining and constrained update orders, limiting its direct applicability to pretrained DLMs. Our token-level analysis reveals pronounced structural locality in DLM inference. Decoding is driven by a small set of prefix-localized active tokens; the influence of distant undecoded context diminishes rapidly, and decoded tokens exhibit stage-wise temporal stability, enabling reuse of intermediate representations except for a brief post-decode transient. Motivated by these observations, we propose \textbf{\placeholder}\footnote{The source code is available at https://github.com/vhicrgit/Window-Diffusion.}, a window-based token pruning and caching method for inference. We maintain a local computation window that slides rightward as denoising progresses, and partition undecoded tokens into: (i) \textit{active tokens} that are computed online, (ii) \textit{buffer tokens} whose KV states are cached and periodically refreshed, and (iii) \textit{far-field tokens} that are pruned outside the window. Computation is restricted to active and buffer tokens within the window, while far-field tokens are omitted at each stage. Experiments on LLaDA and Dream show that, under matched compute budgets, our method achieves up to

99\times

inference speedup while largely preserving generation performance.

Fengrui Zuo, Zhiwei Ke, Yiming Liu +3

School of Computer Science and Technology, University of Science and Technology of China · Suzhou Institute of Advanced Research, University of Science and Technology of China

Diffusion Language Model Inference Token Pruning Diffusion Model Caching Attention Mechanisms Diffusion Model Inference Acceleration

Jan 26, 2026cs.LG

ART for Diffusion Sampling: A Reinforcement Learning Approach to Timestep Schedule

We consider time discretization for score-based diffusion models to generate samples from a learned reverse-time dynamic on a finite grid. Uniform and hand-crafted grids can be suboptimal given a budget on the number of time steps. We introduce Adaptive Reparameterized Time (ART), which controls the clock speed of a reparameterized time variable to redistribute computation along the sampling trajectory while preserving the terminal time, with the objective of minimizing the aggregate Euler discretization error. We derive a randomized companion ART-RL that recasts ART as a continuous-time reinforcement learning problem with Gaussian policies, and prove a two-directional bridge between the two: the deterministic ART optimum lifts to an optimal Gaussian policy, and conversely any optimal Gaussian policy must recover the ART control through its mean. This bridge turns continuous-time actor--critic learning into a principled, rather than heuristic, route to the deterministic timestep optimum. Within the official EDM pipeline, ART-RL improves FID on CIFAR--10 across a wide range of budgets; after one-time offline training, the distilled deterministic schedule transfers without retraining to AFHQv2, FFHQ, and ImageNet at no extra inference cost.

Yilie Huang, Wenpin Tang, Xunyu Zhou

Department of Industrial Engineering and Operations Research, Columbia University, New York, NY 10027, USA. · Department of Industrial Engineering and Operations Research & Data Science Institute, Columbia University, New York, NY 10027, USA.

Diffusion Sampling Adaptive Sampling RL for Diffusion Models Diffusion Timestep Scheduling Diffusion Model Inference Acceleration

Jan 25, 2026cs.LG

Streaming-dLLM: Accelerating Diffusion LLMs via Suffix Pruning and Dynamic Decoding

Diffusion Large Language Models (dLLMs) offer a compelling paradigm for natural language generation, leveraging parallel decoding and bidirectional attention to achieve superior global coherence compared to autoregressive models. While recent works have accelerated inference via KV cache reuse or heuristic decoding, they overlook the intrinsic inefficiencies within the block-wise diffusion process. Specifically, they suffer from spatial redundancy by modeling informative-sparse suffix regions uniformly and temporal inefficiency by applying fixed denoising schedules across all the decoding process. To address this, we propose Streaming-dLLM, a training-free framework that streamlines inference across both spatial and temporal dimensions. Spatially, we introduce attenuation guided suffix modeling to approximate the full context by pruning redundant mask tokens. Temporally, we employ a dynamic confidence aware strategy with an early exit mechanism, allowing the model to skip unnecessary iterations for converged tokens. Extensive experiments show that Streaming-dLLM achieves up to 68.2X speedup while maintaining generation quality, highlighting its effectiveness in diffusion decoding. The code is available at https://github.com/xiaoshideta/Streaming-dLLM.

Zhongyu Xiao, Zhiwei Hao, Jianyuan Guo +4

School of information and Electronics, Beijing Institute of Technology, Beijing · Department of Computer Science, City University of Hong Kong, Hong Kong · School of Computer Science, Wuhan University, Wuhan +2

Adaptive Inference Blockwise Diffusion Language Models Diffusion Language Model Inference Token Pruning Diffusion Language Model Decoding Diffusion Model Inference Acceleration

Jan 20, 2026cs.IT

A Gaussian Perspective for Distributional Discrepancy in Generative Diffusion Models

This paper introduces an analytical approach to quantifying and optimizing the distributional discrepancy in generative diffusion models. For a multivariate Gaussian source, we explicitly derive the closed-form evolution trajectory and the resulting Kullback-Leibler (KL) divergence between the distributions of the source data and the reversely sampled data. Asymptotic analysis via the Euler-Maclaurin expansion characterizes the convergence behavior of this KL divergence, extracting its dominant term as an explicit functional of the noise schedule. Minimizing this dominant term via the calculus of variations yields a noise schedule described by a tangent law, inherently determined by the source covariance spectrum. We further prove that the Gaussian source exhibits an extremal property for the KL divergence among general source distributions with a given covariance. We also utilize the analytical KL divergence as a principled metric to identify efficient time discretization strategies for pretrained diffusion models, and demonstrate via experiments over diverse datasets that the identified strategies consistently outperform established baselines, particularly under constrained function evaluation budgets.

Qiang Sun, H. Vincent Poor, Wenyi Zhang

Department of Electronic Engineering and Information Science, University of Science and Technology of China, Hefei, China · Department of Electrical Engineering, Princeton University, Princeton, NJ, USA

Diffusion Sampling Diffusion Model Inference Acceleration

Jan 14, 2026cs.CV

Mixture of Distributions Matters: Dynamic Sparse Attention for Efficient Video Diffusion Transformers

While Diffusion Transformers (DiTs) have achieved notable progress in video generation, this long-sequence generation task remains constrained by the quadratic complexity inherent to self-attention mechanisms, creating significant barriers to practical deployment. Although sparse attention methods attempt to address this challenge, existing approaches either rely on oversimplified static patterns or require computationally expensive sampling operations to achieve dynamic sparsity, resulting in inaccurate pattern predictions and degraded generation quality. To overcome these limitations, we propose a \underline{\textbf{M}}ixture-\underline{\textbf{O}}f-\underline{\textbf{D}}istribution \textbf{DiT} (\textbf{MOD-DiT}), a novel sampling-free dynamic attention framework that accurately models evolving attention patterns through a two-stage process. First, MOD-DiT leverages prior information from early denoising steps and adopts a {distributed mixing approach} to model an efficient linear approximation model, which is then used to predict mask patterns for a specific denoising interval. Second, an online block masking strategy dynamically applies these predicted masks while maintaining historical sparsity information, eliminating the need for repetitive sampling operations. Extensive evaluations demonstrate consistent acceleration and quality improvements across multiple benchmarks and model architectures, validating MOD-DiT's effectiveness for efficient, high-quality video generation while overcoming the computational limitations of traditional sparse attention approaches.

Yuxi Liu, Yipeng Hu, Zekun Zhang +2

Peking University · University of Electronic Science and Technology of China · University of Science and Technology of China

Video Diffusion Models Diffusion Transformer Sparse Attention Diffusion Model Inference Acceleration

Jan 13, 2026cs.CV

SnapGen++: Unleashing Diffusion Transformers for Efficient High-Fidelity Image Generation on Edge Devices

Recent advances in diffusion transformers (DiTs) have set new standards in image generation, yet remain impractical for on-device deployment due to their high computational and memory costs. In this work, we present an efficient DiT framework tailored for mobile and edge devices that achieves transformer-level generation quality under strict resource constraints. Our design combines three key components. First, we propose a compact DiT architecture with an adaptive global-local sparse attention mechanism that balances global context modeling and local detail preservation. Second, we propose an elastic training framework that jointly optimizes sub-DiTs of varying capacities within a unified supernetwork, allowing a single model to dynamically adjust for efficient inference across different hardware. Finally, we develop Knowledge-Guided Distribution Matching Distillation, a step-distillation pipeline that integrates the DMD objective with knowledge transfer from few-step teacher models, producing high-fidelity and low-latency generation (e.g., 4-step) suitable for real-time on-device use. Together, these contributions enable scalable, efficient, and high-quality diffusion models for deployment on diverse hardware.

Dongting Hu, Aarush Gupta, Magzhan Gabidolla +12

Snap Inc. · University of Melbourne · MBZUAI

Diffusion Transformer Diffusion Model Distillation Few-Step Diffusion Sampling Diffusion Model Inference Acceleration On-Device Inference

Dec 29, 2025cs.CV

Stream-DiffVSR: Low-Latency Streamable Video Super-Resolution via Auto-Regressive Diffusion

Diffusion-based video super-resolution (VSR) methods deliver strong perceptual quality but are often unsuitable for latency-sensitive scenarios due to reliance on future frames and expensive multi-step denoising. We propose Stream-DiffVSR, a causally conditioned diffusion framework for efficient online VSR. Operating strictly on past frames, Stream-DiffVSR integrates a four-step distilled denoiser for fast inference, an Auto-regressive Temporal Guidance (ARTG) module that injects motion-aligned cues during latent denoising, and a lightweight temporal-aware decoder with a Temporal Processor Module (TPM) to enhance detail and temporal coherence. Unlike chunk-wise streaming inference, our strictly frame-by-frame causal design avoids sequence-level waiting, substantially reducing time-to-first-frame and end-to-end latency. Stream-DiffVSR processes 720p frames in 0.328 seconds on an RTX 4090 and consistently outperforms prior diffusion-based baselines. Compared with the online state-of-the-art TMP, it improves perceptual quality (LPIPS +0.095). Compared with prior diffusion-based VSR methods such as MGLD-VSR, it reduces per-frame runtime by over 130x. Moreover, Stream-DiffVSR substantially lowers time-to-first-frame for diffusion-based VSR, reducing initial delay from over 4600 seconds to 0.328 seconds, making diffusion-based VSR markedly more practical for low-latency online and streaming deployment. Project page: https://jamichss.github.io/stream-diffvsr-project-page/

Hau-Shiang Shiu, Chin-Yang Lin, Zhixiang Wang +4

National Yang Ming Chiao Tung University · Shanda AI Research Tokyo · MediaTek Inc.

Video Super-Resolution Diffusion Model Inference Acceleration

Dec 28, 2025cs.LG

ReDiF: Resource-Efficient Few-Step Diffusion Distillation via Reinforcement Learning

Step distillation accelerates diffusion sampling by training a few-step student to imitate a many-step teacher, but distillation itself remains expensive. Typically, this requires thousands of GPU-hours and a large pre-generated trajectory dataset. We introduce ReDiF, which casts step distillation as terminal-reward policy optimization rather than step-wise regression. The student is optimized against a reward computed on the terminal sample, measuring alignment with the teacher's output, instead of matching the teacher's intermediate trajectory under a reconstruction or consistency loss. Because the reward need not be differentiable or trajectory-aligned, ReDiF admits non-differentiable objectives, multi-objective combinations, and preferences the teacher does not express, while exploration lets the student find sampling paths matched to its own step schedule. ReDiF converges in 400 policy updates with 3200 rollouts on a single A100 GPU with 1,000 noise-class pairs and no paired dataset: about 4 GPU-hours, against roughly 336 A100-hours reported for DMD2 on the same EDM teacher. At 8 steps on ImageNet-64, it achieves an FID 3.64 points better than the strongest retrained distillation baseline in the low-training regime under the same single-GPU budget. The formulation is also orthogonal to existing distillation objectives: added to the DMD2 loss, it further improves DMD2's FID by 4.4.

Amirhossein Tighkhorshid, Zahra Dehghanian, Hamid R. Rabiee

Computer Engineering Sharif University of Technology

Diffusion Distillation RL for Diffusion Models Teacher-Student Learning Diffusion Model Inference Acceleration

Dec 16, 2025cs.CV

Sparse-LaViDa: Sparse Multimodal Discrete Diffusion Language Models

Masked Discrete Diffusion Models (MDMs) have achieved strong performance across a wide range of multimodal tasks, including image understanding, generation, and editing. However, their inference speed remains suboptimal due to the need to repeatedly process redundant masked tokens at every sampling step. In this work, we propose Sparse-LaViDa, a novel modeling framework that dynamically truncates unnecessary masked tokens at each inference step to accelerate MDM sampling. To preserve generation quality, we introduce specialized register tokens that serve as compact representations for the truncated tokens. Furthermore, to ensure consistency between training and inference, we design a specialized attention mask that faithfully matches the truncated sampling procedure during training. Built upon the state-of-the-art unified MDM LaViDa-O, Sparse-LaViDa achieves up to a 2x speedup across diverse tasks including text-to-image generation, image editing, and mathematical reasoning, while maintaining generation quality.

Shufan Li, Jiuxiang Gu, Kangning Liu +4

Adobe · UCLA

Masked Diffusion Models Diffusion Language Model Inference Token Pruning Diffusion Model Inference Acceleration Diffusion Models Multimodal Diffusion Models

Dec 9, 2025cs.CV

Early Estimation of Language to Latent Alignment in Diffusion Models

Conditional diffusion models frequently suffer from language-image misalignments. Due to the ambiguity of intermediate noise corrupted latents, assessing prompt adherence currently requires completing the entire sampling trajectory. This late-stage evaluation incurs even higher computational costs during test-time scaling strategies, such as Best-of-N (BoN) sampling, as all misaligned trajectories must finish generation before being discarded. To tackle this, we propose NoisyCLIP, a noise-aware twin-tower model that enables early language-to-latent alignment estimation. By learning a vision encoder on noise-corrupted latents, we allow the model to "see" through the ambiguity of intermediate diffusion steps. To facilitate this training, we investigate noise-data augmentation sampling strategies and introduce two new benchmark datasets: Noisy-Conceptual-Captions and Noisy-GenAI-Bench. When applied as an early-stopping criterion for BoN, NoisyCLIP at half cost matches or beats frozen CLIP at full cost. Ultimately, this transforms alignment assessment from an expensive final check into a continuous monitoring tool, drastically reducing compute costs without sacrificing semantic fidelity.

Vasco Ramos, Regev Cohen, Idan Szpektor +1

NOVALINCS, NOVA University of Lisbon, Portugal · Google Research

Diffusion Model Alignment Vision-Language Alignment Conditional Diffusion Models Diffusion Model Inference Acceleration

Nov 27, 2025cs.CV

Z-Image: An Efficient Image Generation Foundation Model with Single-Stream Diffusion Transformer

The landscape of high-performance image generation models is currently dominated by proprietary systems, such as Nano Banana Pro and Seedream 4.0. Leading open-source alternatives, including Qwen-Image, Hunyuan-Image-3.0 and FLUX.2, are characterized by massive parameter counts (20B to 80B), making them impractical for inference, and fine-tuning on consumer-grade hardware. To address this gap, we propose Z-Image, an efficient 6B-parameter foundation generative model built upon a Scalable Single-Stream Diffusion Transformer (S3-DiT) architecture that challenges the "scale-at-all-costs" paradigm. By systematically optimizing the entire model lifecycle -- from a curated data infrastructure to a streamlined training curriculum -- we complete the full training workflow in just 314K H800 GPU hours (approx. $630K). Our few-step distillation scheme with reward post-training further yields Z-Image-Turbo, offering both sub-second inference latency on an enterprise-grade H800 GPU and compatibility with consumer-grade hardware (<16GB VRAM). Additionally, our omni-pre-training paradigm also enables efficient training of Z-Image-Edit, an editing model with impressive instruction-following capabilities. Both qualitative and quantitative experiments demonstrate that our model achieves performance comparable to or surpassing that of leading competitors across various dimensions. Most notably, Z-Image exhibits exceptional capabilities in photorealistic image generation and bilingual text rendering, delivering results that rival top-tier commercial models, thereby demonstrating that state-of-the-art results are achievable with significantly reduced computational overhead. We publicly release our code, weights, and online demo to foster the development of accessible, budget-friendly, yet state-of-the-art generative models.

Z-Image Team, Huanqia Cai, Sihan Cao +21

Z-Image Team, Alibaba Group

Diffusion Transformer Diffusion Model Inference Acceleration T2I Diffusion Models

Nov 21, 2025cs.AI

DAPS++: Rethinking Diffusion Inverse Problems with Decoupled Posterior Annealing

From a Bayesian perspective, score-based diffusion solves inverse problems through joint inference, embedding the likelihood with the prior to guide the sampling process. However, this formulation fails to explain its practical behavior: the prior offers limited guidance, while reconstruction is largely driven by the measurement-consistency term, leading to an inference process that is effectively decoupled from the diffusion dynamics. We show that the diffusion prior in these solvers functions primarily as a warm initializer that places estimates near the data manifold, while reconstruction is driven almost entirely by measurement consistency. Based on this observation, we introduce \textbf{DAPS++}, which fully decouples diffusion-based initialization from likelihood-driven refinement, allowing the likelihood term to guide inference more directly while maintaining numerical stability and providing insight into why unified diffusion trajectories remain effective in practice. By requiring fewer function evaluations (NFEs) and measurement-optimization steps, \textbf{DAPS++} achieves high computational efficiency and robust reconstruction performance across diverse image restoration tasks.

Hao Chen, Renzheng Zhang, Scott S. Howard

University of Notre Dame, Notre Dame, IN 46556, USA

Diffusion Models for Image Restoration Bayesian Inverse Problems Diffusion-Based Inverse Problems Diffusion Posterior Sampling Diffusion Model Inference Acceleration Image Inverse Problems

Oct 18, 2025cs.CV

UltraImageGen: Efficient Ultra-High-Resolution Image Generation with Hierarchical Local Attention

Ultra-high-resolution text-to-image generation is increasingly vital for applications requiring fine-grained textures and global structural fidelity, yet state-of-the-art text-to-image diffusion models such as FLUX and SD3 remain confined to sub 2MP (<

1K\times2K

) resolutions due to the quadratic complexity of attention mechanisms and the scarcity of high-quality high-resolution training data. We present UltraImageGen, a novel framework that introduces hierarchical local attention with low-resolution global guidance, enabling efficient, scalable, and semantically coherent image synthesis at ultra-high resolutions. Specifically, high-resolution latents are divided into hardware aligned fixed-size local windows to reduce attention complexity from quadratic to near-linear, while a low-resolution latent equipped with scaled positional embeddings injects global semantics as an anchor. A lightweight LoRA adaptation bridges global and local pathways during denoising, ensuring consistency across structure and detail. To maximize inference efficiency and achieve scalable ultra-high-resolution generation, we repermute token sequence in window-first order, so that the GPU-friendly dense local blocks in attention calculation equals to the fixed-size local window in 2D regardless of resolution. Together ourwork reliably scales the pretrained model to resolutions higher than

8K

with more than

10\times

speed up and significantly lower memory usage. Extensive experiments demonstrate that ourwork achieves superior quality while maintaining computational efficiency, establishing a practical paradigm for advancing ultra-high-resolution image generation.

Yuyao Zhang, Yu-Wing Tai

Dartmouth College, USA

Image Generation Diffusion Model Inference Acceleration T2I Diffusion Models

Sep 29, 2025cs.CV

DC-Gen: Post-Training Diffusion Acceleration with Deeply Compressed Latent Space

Existing text-to-image diffusion models excel at generating high-quality images, but face significant efficiency challenges when scaled to high resolutions, like 4K image generation. While previous research accelerates diffusion models in various aspects, it seldom handles the inherent redundancy within the latent space. To bridge this gap, this paper introduces DC-Gen, a general framework that accelerates text-to-image diffusion models by leveraging a deeply compressed latent space. Rather than a costly training-from-scratch approach, DC-Gen uses an efficient post-training pipeline to preserve the quality of the base model. A key challenge in this paradigm is the representation gap between the base model's latent space and a deeply compressed latent space, which can lead to instability during direct fine-tuning. To overcome this, DC-Gen first bridges the representation gap with a lightweight embedding alignment training. Once the latent embeddings are aligned, only a small amount of LoRA fine-tuning is needed to unlock the base model's inherent generation quality. We verify DC-Gen's effectiveness on SANA and FLUX.1-Krea. The resulting DC-Gen-SANA and DC-Gen-FLUX models achieve quality comparable to their base models but with a significant speedup. Specifically, DC-Gen-FLUX reduces the latency of 4K image generation by 53x on the NVIDIA H100 GPU. When combined with NVFP4 SVDQuant, DC-Gen-FLUX generates a 4K image in just 3.5 seconds on a single NVIDIA 5090 GPU, achieving a total latency reduction of 138x compared to the base FLUX.1-Krea model. Code: https://github.com/dc-ai-projects/DC-Gen.

Wenkun He, Yuchao Gu, Junyu Chen +11

Diffusion Model Fine-Tuning Diffusion Model Inference Acceleration T2I Diffusion Models

Sep 25, 2025cs.CL

Enabling Approximate Joint Sampling in Diffusion LMs

In autoregressive language models, each token is sampled by conditioning on all the past tokens; the overall string has thus been sampled from the correct underlying joint distribution represented by the model. In contrast, masked diffusion language models generate text by unmasking tokens out of order and potentially in parallel. Generating an overall string sampled from the correct underlying joint distribution would (again) require exactly one token unmasking in every full-model forward pass. The more tokens unmasked in parallel, the further away the string is from the true joint; this can be seen in the resulting drop in accuracy (but, increase in speed). In this paper we devise a way to {\em approximately} sample multiple tokens from the joint distribution in a single full-model forward pass; we do so by developing a new lightweight single-layer ``sampler" on top of an existing large diffusion LM. One forward pass of the full model can now be followed by multiple forward passes of only this sampler layer, to yield multiple unmasked tokens. Our sampler is trained to mimic exact joint sampling from the (frozen) full model. We show the effectiveness of our approximate joint sampling for both pretrained-only (Dream-7B-Base, Llada-7B-Base) and instruction-tuned (Dream-7B-Instruct, Dream-7B-Coder) models on language modeling and math & coding tasks. When four tokens are unmasked for each full-model denoising step, our sampling algorithm achieves a MAUVE score of 0.87 (vs marginal baseline of 0.31) with respect to the true joint distribution.

Parikshit Bansal, Sujay Sanghavi

UT Austin

Masked Diffusion Models Diffusion Language Model Decoding Diffusion Language Models Diffusion Model Inference Acceleration