Glance Forcing: Making Bidirectional Diffusion Autoregressive at a Glance
Authors: Zhuobai Dong, Junchao Yi, Hu Jian Guo, Linjie Li, Rui Zhao, Alex Jinpeng Wang
Organizations: Wuhan University · University of Electronic Science and Technology of China · Sun Yat-sen University · University of Washington · National University of Singapore · Central South University
Autoregressive generation underlies real-time, causal video synthesis and is the backbone of video world models. Today's dominant approach distills a bidirectional diffusion model into a causal few-step student, but existing pipelines make this conversion prohibitively expensive, requiring offline Probability Flow ODE} (PF-ODE) trajectories that cost tens of thousands of GPU-hours and terabytes of storage. We show it can be made dramatically more efficient. Pushing this efficiency to its extreme, we reduce trajectory supervision to a single sample, the strictest setting possible, and find that it causes severe quality collapse. We trace this to the teacher's trajectory geometry: the high-noise regime, governing semantic layout and motion, and the low-noise regime, governing fine texture exhibit fundamentally different curvature and optimization behavior, so a single monolithic student cannot jointly fit both, causing severe representation interference. Motivated by this, we propose Glance Forcing, a lightweight distillation framework that decouples the denoising trajectory into two complementary slow and fast experts, each guided by a tailored, stage-specific objective, while keeping the pretrained teacher entirely frozen. This decoupling further reveals a general mechanism underlying few-step distillation: PF-ODE distillation mainly straightens the few-step trajectory geometry, while the subsequent distribution-matching stage independently governs distributional quality. Using only one PF-ODE trajectory, Glance Forcing already surpasses the fully-trained state-of-the-art baseline, raising the VBench Total score from 84.04 to 84.22 while cutting trajectory-collection cost from thousands of GPU-hours to about one A100-GPU-hour, with markedly larger gains in dynamic degree, VisionReward, and instruction-following, and consistent preference from human raters over strong baselines.
Figures & tables
Figure 1: Overall framework of our Glance Forcing and the comparison with existing methods. (a) Causal Forcing (CF) fixes Self Forcing (SF)’s architecture gap but remains costly; our Glance Forcing (GF) is efficient, and superior. (b) Our GF reduces training cost by 11600×, requires only 1 extra sample curation, and achieves higher VBench scores than CF.
Figure 2: Motivation and Trajectory Analysis. (a) VBench performance drops significantly for both forcing methods when trained on a single sample on the ODE stage. (b) Relative trajectory-curvature profiles show high-noise concentration for Wan video generation. (c) Distillation loss exhibit significant differences between samples with high and low noise levels.
Figure 3: Overview of the Slow-Fast paradigm. We introduce a chunk-wise autoregressive diffusion framework, where each chunk is progressively denoised with stage-specific LoRA adapters. Slow-LoRA focuses on coarse structure generation in early stages, while Fast-LoRA refines details in later stages. Expert-specific objectives further enhance temporal consistency and spatial fidelity.
Figure 4: Comparison of cosine similarity trends and trajectory straightness across different models.
Figure 5: Comparison of visual quality across denoiser variants trained on different noise levels . Decoupling distillation into semantic and detail learning with two expert denoisers and combining them at inference yields the best quantitative visual results.
Model
Data
Thru. ↑
Latency ↓
Total ↑
Quality ↑
Semantic ↑
Dyn. ↑
Vision. ↑
Instruct. ↑
Bidirectional Video Diffusion Models
Wan2.1-1.3B
-
0.78
103
83.37
84.30
79.65
61
5.275
42
LTX-1.9B
-
8.98
13.5
79.83
81.88
71.62
46
-6.218
-38
Autoregressive Video Diffusion Models
SkyReels-V2
-
0.49
112
81.97
83.96
74.01
37
3.584
32
MAGI-1
-
0.19
282
78.88
81.67
67.72
42
0.773
8
Table 1: Comparisons with existing methods. Our method achieves generation quality comparable to or better than previous SOTA methods. Data, Dyn., Vision., and Instruct. denote Training Data, Dynamic Degree, VisionReward, and Instruction Following, respectively. * indicates the same set of models across stages. Throughput and latency are measured in FPS and seconds on H100 GPU.
Figure 6: Qualitative comparisons with existing methods. Our method achieves competitive visual quality than state-of-the-art distilled autoregressive video models.
Table 8
Figure 7: Ablation study Performance comparison.
Figure 8: Qualitive results from one-sample training setting.
Table 11
Appendix figures & tables4 assets
Supplementary material from the paper’s appendix.
Appendix
Figure 9: User Study platform.
Figure 10: Impact of different κ .
Figure 11: Qualitative comparisons with existing methods for Initialized stage.
Figure 12: Qualitative comparisons with existing methods for distilled stage.
Distilling pretrained video diffusion models into few-step autoregressive (AR) generators typically combines student initialization with distribution matching distillation (DMD). An intermediate AR teacher can provide causally compatible targets for ODE initialization. However, causally compatible targets alone do not ensure high-quality generation under a short sampling schedule. We analyze the posterior-mean predictions favored by direct latent regression and empirically find that causal ODE initialization alleviates blur but still leaves fine detail unresolved. To address this limitation, we propose Representation Forcing, which learns few-step causal initialization through representation-space supervision rather than intermediate-teacher trajectory regression. The student is trained with teacher forcing by comparing clean predictions and training targets through normalized features of a frozen video diffusion transformer. This phase reuses the synthetic videos generated by a bidirectional model for AR teacher training and replaces both teacher training and causal ODE distillation, without generating additional causal ODE trajectories. The subsequent asymmetric DMD procedure remains unchanged. Empirical results show that, at four sampling steps, RepForcing better preserves generation quality than intermediate-teacher-based initialization before DMD. After DMD, our four-step chunk-wise model improves VBench Total from 84.04 to 84.66 (+0.62 points) and VisionReward from 6.326 to 10.074 (+59.25%) over the reported causal-ODE-based baseline.
Real-time interactive video generation requires low-latency, streaming, and controllable rollout. Existing autoregressive (AR) diffusion distillation methods have achieved strong results in the chunk-wise 4-step regime by distilling bidirectional base models into few-step AR students, but they remain limited by coarse response granularity and non-negligible sampling latency. In this paper, we study a more aggressive setting: frame-wise autoregression with only 1--2 sampling steps. In this regime, we identify the initialization of a few-step AR student as the key bottleneck: existing strategies are either target-misaligned, incapable of few-step generation, or too costly to scale. We propose Causal Forcing++, a principled and scalable pipeline that uses causal consistency distillation (causal CD) for few-step AR initialization. The core idea is that causal CD learns the same AR-conditional flow map as causal ODE distillation, but obtains supervision from a single online teacher ODE step between adjacent timesteps, avoiding the need to precompute and store full PF-ODE trajectories. This makes the initialization both more efficient and easier to optimize. The resulting pipeline, \ours, surpasses the SOTA 4-step chunk-wise Causal Forcing under the frame-wise 2-step setting by 0.1 in VBench Total, 0.3 in VBench Quality, and 0.335 in VisionReward, while reducing first-frame latency by 50% and Stage 2 training cost by \sim$$4\times. We further extend the pipeline to action-conditioned world model generation in the spirit of Genie3. Project Page: https://github.com/thu-ml/Causal-Forcing and https://github.com/shengshu-ai/minWM .
Min Zhao, Hongzhou Zhu, Kaiwen Zheng +6
Tsinghua University · ShengShu · Renmin University of China
Few-step autoregressive video generation commonly relies on Distribution Matching Distillation (DMD), requiring a bidirectional diffusion teacher and an online fake-score model. We instead learn the rollout distribution directly from reference videos, eliminating both score models during post-training. Our framework minimizes maximum mean discrepancy (MMD) in frozen self-supervised video representation spaces, using a hybrid Nyström--Monte Carlo estimator to balance approximation bias and sampling variance. Memory-efficient replay and gradient subsampling make this objective practical. Using the same architecture and initialization as Self-Forcing, our 1.3B model improves the VBench Total score from 83.80 to 84.64 while retaining 17 FPS. Removing auxiliary score models also enables 14B post-training on eight H200 GPUs. Beyond distillation, learning from reference videos enables the acquisition of new visual styles, semantic concepts, and spatial priors without a target-specific diffusion teacher.
Chi Zhang, Yueyi Liu, Shi Haoyang +5
College of AI, Tsinghua University · IAIR, Xi’an Jiaotong University · Xianghui Academy, Fudan University +1