Pith. sign in

REVIEW 12 cited by

Applying Guidance in a Limited Interval Improves Sample and Distribution Quality in Diffusion Models

Not yet reviewed by Pith; the record is open.

Re-run · record.json Download PDF Read on arXiv ↗

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2404.07724 v2 pith:HJBHH3HI submitted 2024-04-11 cs.CV cs.AIcs.LGcs.NEstat.ML

Applying Guidance in a Limited Interval Improves Sample and Distribution Quality in Diffusion Models

Tuomas Kynk\"a\"anniemi , Miika Aittala , Tero Karras , Samuli Laine , Timo Aila , Jaakko Lehtinen This is my paper
classification cs.CV cs.AIcs.LGcs.NEstat.ML
keywords guidancediffusionintervallevelsmodelsnoisebeneficialchain
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
read the original abstract

Guidance is a crucial technique for extracting the best performance out of image-generating diffusion models. Traditionally, a constant guidance weight has been applied throughout the sampling chain of an image. We show that guidance is clearly harmful toward the beginning of the chain (high noise levels), largely unnecessary toward the end (low noise levels), and only beneficial in the middle. We thus restrict it to a specific range of noise levels, improving both the inference speed and result quality. This limited guidance interval improves the record FID in ImageNet-512 significantly, from 1.81 to 1.40. We show that it is quantitatively and qualitatively beneficial across different sampler parameters, network architectures, and datasets, including the large-scale setting of Stable Diffusion XL. We thus suggest exposing the guidance interval as a hyperparameter in all diffusion models that use guidance.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 12 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Guidance Breaks the Fitted Operator: A Terminal-Fitted Repair for Classifier-Free Guidance

    cs.LG 2026-07 conditional novelty 7.0

    Replacing CFG's w(r-1) coefficient with r^(1+w)-r removes a sigma_min-divergent residual blow-up on a Gaussian calibration model and stabilizes high-guidance diffusion sampling at zero extra NFE.

  2. Factored Classifier-Free Guidance

    cs.CV 2025-06 unverdicted novelty 7.0

    Factored Classifier-Free Guidance enables per-attribute control in classifier-free guidance for diffusion models to produce more sound counterfactuals.

  3. Unified Audio Intelligence Without Regressing on Text Intelligence

    cs.CL 2026-07 conditional novelty 6.0

    Audex unifies audio understanding and generation on a strong text MoE backbone with multi-stage SFT plus text-only Cascade RL, matching open SOTA audio scores while mostly retaining text capability.

  4. DiffusionBench: On Holistic Evaluation of Diffusion Transformers

    cs.CV 2026-06 conditional novelty 6.0

    NanoGen unifies DiT training on ImageNet and T2I, reveals negative Pearson correlations (-0.377 to -0.580) in method rankings across metrics from 21 models, and motivates DiffusionBench for holistic evaluation.

  5. Information-Theoretic Classifier-Free Guidance with Adaptive Schedule Optimization

    cs.LG 2026-06 unverdicted novelty 6.0

    An information-theoretic method optimizes CFG schedules in diffusion models by targeting a reference consistency-coverage trade-off via trajectory-level estimates from samples.

  6. Improved Baselines with Representation Autoencoders

    cs.CV 2026-05 conditional novelty 6.0

    RAE v2 reaches gFID 1.06 on ImageNet-256 in 80 epochs by combining multi-layer encoder sums, complementary REPA targets, and free guidance via output reparameterization.

  7. VAGS: Velocity Adaptive Guidance Scale for Image Editing and Generation

    cs.CV 2026-05 accept novelty 6.0

    VAGS adapts the CFG scale at each ODE step using velocity alignment signals to raise structural fidelity in editing and sample quality in generation over fixed-scale baselines.

  8. PixelGen: Improving Pixel Diffusion with Perceptual Supervision

    cs.CV 2026-02 accept novelty 6.0

    PixelGen augments pixel diffusion with gated perceptual supervision to reach FID 5.11 on ImageNet-256 and GenEval 0.79 in text-to-image, narrowing the gap to latent methods without VAEs.

  9. DeCo: Frequency-Decoupled Pixel Diffusion for End-to-End Image Generation

    cs.CV 2025-11 conditional novelty 6.0

    DeCo decouples high- and low-frequency generation in pixel diffusion via a DiT plus lightweight decoder and a frequency-aware flow-matching loss, reaching FID 1.62 at 256x256 and 2.22 at 512x512 on ImageNet while clos...

  10. Representation Alignment for Generation: Training Diffusion Transformers Is Easier Than You Think

    cs.CV 2024-10 unverdicted novelty 6.0

    Aligning noisy hidden states in diffusion transformers to clean features from pretrained visual encoders speeds up training over 17x and reaches FID 1.42.

  11. Unified Audio Intelligence Without Regressing on Text Intelligence

    cs.CL 2026-07 conditional novelty 5.0

    A unified 30B MoE audio-text LLM achieves state-of-the-art audio understanding, generation, and speech tasks while preserving text reasoning comparable to its text-only backbone.

  12. Elucidating Representation Degradation Problem in Diffusion Model Training

    cs.LG 2026-05 unverdicted novelty 4.0

    Diffusion models suffer representation degradation at high noise due to recoverability mismatch; ERD mitigates this by dynamic optimization reallocation, accelerating convergence across backbones.