LongLive‑Plug is a once‑for‑all distillation framework that learns reusable LoRA adapters on a base video diffusion model, enabling training‑free, plug‑and‑play deployment to a wide range of downstream models. These adapters provide single‑pass classifier‑free guidance, few‑step sampling, and long‑context error correction for autoregressive generation, and remain effective even when downstream models add conditioning branches or expand output channels. The authors demonstrate that the approach works on 54 downstream models across three backbone families and eight task categories, including world modeling, robotics, editing, and multimodal generation.
By Shuai Yang, Luozhou Wang, Wei Huang, ZhiFei Chen, Bohan Zhang, Xiao Fu, Qianli Ma, Chen-Hsuan Lin, Weian Mao, Bryan Chu, Song Han, Yukang Chen
arXiv:2607. 07665v1 Announce Type: new Abstract: Classifier-free guidance (CFG) is the standard way to strengthen class-conditioning in diffusion and flow-matching samplers, yet at large guidance it oversaturates and destabilizes, symptoms practitioners suppress with more steps or limited-interval schedules.
By Shiheng Zhang
Classifier-free guidance (CFG) is the standard way to strengthen class-conditioning in diffusion and flow-matching samplers, yet at large guidance it oversaturates and destabilizes, symptoms practitioners suppress with more steps or limited-interval schedules. We analyze CFG through an asymptotic-preserving, numerical-analysis lens.
arXiv:2602. 20360v2 Announce Type: replace Abstract: Flow-based generative methods offer a simple and effective framework for high-fidelity generation, yet pretrained flow models are rarely used in their vanilla conditional form: in image generation, samples without guidance often appear diffuse and lack fine-grained detail.
By Runlong Liao, Jian Yu, Baiyu Su, Chi Zhang, Lizhang Chen, Qiang Liu
TrajCFG introduces a training‑free, token‑wise adaptation of classifier‑free guidance for flow‑matching models, leveraging historical changes in guided velocity direction to assign bounded scales. By mapping directional feedback to stronger guidance for larger changes and weaker guidance for consistent directions, it fuses adjacent‑step and interval‑normalized longer‑span signals without extra model evaluations. Experiments across text‑to‑video, image‑to‑video, text‑to‑audio‑video, text‑to‑image, and image‑to‑3D tasks demonstrate consistent improvements over base samplers, notably raising VBench Total scores by 0.30‑0.75 points on three backbones for text‑to‑video generation.
By Jingyu Liu, Hao Jiang, Yikang Ding, Huafeng Shi, Xu Yang, Xiaoqiang Liu, Guoxin Zhang, Pengfei Wan, Kun Gai
arXiv:2608.29107v1 Announce Type: new
Abstract: While modern generative models excel at modeling complex data, precise inference-time control in conditional generation remains a critical challenge. C...
By Avishag Nevo, Tamir Hazan
arXiv:2610.26025v2 Announce Type: replace-cross
Abstract: Diffusion Transformers (DiTs) deliver high-fidelity image and video generation, but inference requires many sequential model evaluations. Two...
By Shuaiqi Cheng, Jiahao Zhang, Boyuan Gu, Yuxi Chen, Haiying Xu, Zexin Lu, Huiyu Zhou, Xuming Hu
The paper introduces Global Transport (GT), a class‑agnostic optimal‑transport coupling that can be computed without class labels. GT associates different conditions with distinct regions of the source noise, which degrades performance when used without guidance but consistently improves generation when combined with classifier‑free guidance across various domains, model scales, and sampling budgets. The authors argue that coupling design should be evaluated under guided flow conditions rather than unguided generation, and demonstrate GT’s benefits on both discrete class‑conditioned and continuous text‑conditioned image generation.
By Katarina Petrovi\'c, Zander W. Blasingame, Danyal Rehman, \.Ismail \.Ilkan Ceylan, Michael Bronstein, Stephen Y. Zhang, Lazar Atanackovic, Alexander Tong
The paper introduces TRACK, a training‑free trajectory routing method that accelerates video diffusion by selectively switching between large and small models during denoising steps. A calibration process generates a disagreement score map, guiding the selection of the appropriate model at each step to maintain quality while reducing computational cost. Experiments on Wan 2.1, Cosmos 3, TurboDiffusion, and FastVideo show speedups ranging from 1.95× to 2.73× with comparable quality and diversity.
By Mustafa Munir, Huy Vu, Shreyas Misra, Rohit Jena, Sajad Norouzi, Ali Taghibakhshi, Anis Ahmad, Anjul Patney, Pavlo Molchanov, Nima Tajbakhsh
arXiv:2607. 09133v1 Announce Type: cross Abstract: While large-scale text-to-image generative models have achieved unprecedented visual performance, their inherent reliance on multi-step iterative solvers incurs severe inference latency.
By Yiting Wang, Jingyi Zhang, Wenhu Zhang, Ke Chao, Yves Liang, Kun Cheng, Kang Zhao
arXiv:2610.25844v2 Announce Type: replace-cross
Abstract: Training high-quality diffusion models is computationally expensive, and inference-time guidance often adds further sampling cost. In this wo...
By Yanjin He, Cuifeng Shen, Wanzhou Lei, Maohua Li, Tao Lan, Kan Liu, Hanlin Tang, Qing Qu
The paper introduces a lightweight correction framework that improves instruction following in few-step text-to-image generators without altering the generator’s parameters or increasing forward-pass count. The framework adds a small corrector module (<20 M parameters) that learns residual velocity adjustments at selected sampling steps, using On‑Trajectory Flow‑Bridge Supervision to target the actual states reached during corrected generation. Experiments show a +17.60‑point NetWin over a frozen baseline with only a 0.84% increase in end‑to‑end latency.
By Yating Gao, Zhenlong Zhu, Wen Zihao, Lenan Zhang, Jiexiong Wang, Yang Yang, Guanjun Jiang, Yixiong Zou, Takuya Maekawa