Guidance Is a Field, Not a Scalar: Interval-Integrated Classifier-Free Guidance for One-Step Video Generation

Zijie Meng, Xixin CaoarXiv 2026 · arXiv

TL;DR

AGAVE (Adaptive Guidance-Aware Velocity Ensembling) transforms classifier-free guidance from a single, rigid scalar into a dynamic field that varies across time, space, and video frames. By mathematically "baking" this guidance schedule directly into the training target of a one-step model, the authors achieve the quality of complex, multi-step adaptive sampling in a single function evaluation with zero extra inference cost.

What Problem Does This Paper Address?

Diffusion and flow-matching models are the current gold standard for video generation, but they are notoriously slow, requiring dozens of network evaluations to produce a single clip. "One-step" generators like MeanFlow aim to solve this by jumping directly from noise to data in a single step.

However, these fast models have a "guidance problem." Classifier-Free Guidance (CFG)—the "knob" used to trade off diversity for prompt alignment—is usually treated as a single constant scalar. But research shows that guidance shouldn't be constant: it should be low at the start of denoising (to avoid errors), peak in the middle, and vary spatially based on where the action is.

Multi-step models can easily change the guidance scale at every step. One-step models, by definition, have no "steps" to schedule over, forcing them to use a crude, fixed value that often leads to over-saturated colors or "collapsed" motion in videos.

Methods

The authors resolve this tension with a clever mathematical observation. MeanFlow models don't just predict a single point; they learn an average velocity over a noise interval.

The Interval-Integrated CFG (I²-CFG) Identity

The core contribution is Theorem 1. The authors prove that because the MeanFlow target is an integral, you can place a time-varying guidance function inside that integral. This allows a single network evaluation to "compile" the effect of an entire multi-step guidance schedule.

Figure 1. Overview of AGAVE. A single MeanFlow DiT uθ emits a conditional/unconditional velocity pair whose guidance gap ∆v is modulated along three axes – a learnable Beta schedule over noise level τ, a free motion-saliency map M over space, and a learnable frame-position profile ρ(f) – forming a continuous guidance field ω(zτ, τ, f). The interval-integrated CFG operator (Thm. 1) absorbs this field into the average-velocity target, compiling a full time-, space- and frame-varying schedule into one evaluation. The schedule parameters ϕ are learned by distilling a multi-step adaptively guided teacher; inference remains strictly 1-NFE.

AGAVE: Three Axes of Guidance

The researchers instantiate this theory as AGAVE, which modulates guidance across three dimensions:

  1. Noise-time: Uses a learnable Beta-shaped schedule. This allows the model to "rediscover" the optimal strategy of peaking guidance in the middle of the process.
  2. Space (Motion-Saliency): Automatically amplifies guidance in regions with coherent motion (foreground) while keeping it lower for static backgrounds.
  3. Frame (Position Asymmetry): Learns a profile that treats interior frames differently than boundary frames to improve temporal consistency.
The learnable Beta schedule integrates the entire guidance path into one evaluation.

To learn these parameters, they use adaptive-guidance distillation, where a multi-step "teacher" model provides a reference for what a high-quality, guided sample should look like.

Datasets

The authors evaluated AGAVE on several standard video generation benchmarks:

  • VBench: A comprehensive suite for video quality, including motion, consistency, and aesthetics.
  • UCF-101: Used for measuring FVD (Fréchet Video Distance).
  • MSR-VTT: Another standard dataset for text-to-video evaluation.
  • ImageNet 256²: Used for a proof-of-concept image generation task.

Key Insights and Ideas

  • Guidance is a Field: The most important shift is moving from a scalar $\omega$ to a field $\omega(z_\tau, \tau, f)$. This acknowledges that different parts of a video need different amounts of "push" from the prompt.
  • Amortization of Control: Instead of paying the computational price of a schedule at inference time, AGAVE pays it once during training.
  • The "No Discretization Error" Advantage: Surprisingly, a one-step model using I²-CFG can actually outperform a multi-step model. Multi-step models only approximate a guidance schedule through discrete steps (incurring error), whereas the one-step model integrates the schedule exactly.
Figure 4. Compilation gap. A K-step adaptive sampler only reaches the VBench total that AGAVE attains at K=1 (dashed) as K grows, visualizing the Θ(1/K) error of Prop. 1.

Key Results

  • State-of-the-Art: AGAVE set a new SOTA for one-step video generation on VBench, surpassing previous leaders like iMF and rCM.
  • Recovering Motion: It specifically improved "Dynamic Degree" (motion) by +4.7 points on the 14B parameter model, fixing the "static video" problem common in distilled models.
  • Efficiency: It matches the quality of a 50-step teacher model in just 1 step, with zero extra latency and fewer than ten additional learnable scalars.
  • Learned Behavior: Without manual tuning, the model learned a bell-shaped temporal schedule that naturally vanishes at high noise—validating existing "zero-init" heuristics used in multi-step sampling.
The learned noise-time schedule naturally forms a bell shape, peaking in the mid-denoising phase.

Limitations and Future Work

  • Distillation Dependency: While the schedule parameters are few, AGAVE currently requires a multi-step teacher to provide the distillation signal.
  • Background Consistency: The authors note that while AGAVE improves motion, some baselines score slightly higher on background consistency simply because they produce nearly static videos—a trade-off AGAVE intentionally breaks to provide more dynamic content.
  • Broader Control: The "Interval-Integrated" framework could potentially be used for other types of inference-time controls beyond just CFG, such as structural guidance or style transfer.

Summary updated

Abstract

One-step generators built on MeanFlow synthesize a video in a single function evaluation, yet inherit a rigid treatment of classifier-free guidance (CFG): the guidance scale is frozen during training or, at best, exposed as a single scalar. This conflicts with mounting evidence that the optimal guidance is not a constant but a field that varies with the noise level, with spatial content, and, for video, with frame position. Every existing remedy presupposes a multi-step sampler along which guidance can be re-scheduled--an assumption one-step models violate by construction. We resolve this tension through a single observation: because the MeanFlow target is an integral over a noise interval, an arbitrary guidance field can be absorbed into the target itself and compiled into one evaluation. We formalize this as the interval-integrated CFG identity, prove that the compiled guidance carries no schedule-discretization error and can thus match or exceed any finite multi-step sampler, and instantiate it as AGAVE, which equips the average-velocity field with a learnable temporal schedule, a free motion-saliency spatial modulation, and a frame-position asymmetry. Adding fewer than ten scalars at zero inference cost, AGAVE surpasses all one-step baselines on VBench, UCF-101 and MSR-VTT while matching multi-step guided sampling. Project repository: https://github.com/YMlinfeng/AGAVE.
BibTeX
@misc{meng2026guidance,
  author = {Zijie Meng and Xixin Cao},
  title = {{Guidance Is a Field, Not a Scalar: Interval-Integrated Classifier-Free Guidance for One-Step Video Generation}},
  year = {2026},
  eprint = {2611.07566},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2611.07566},
}
Figure