Saddle-Free Guidance: Improved On-Manifold Sampling without Labels or Additional Training
- URL: http://arxiv.org/abs/2511.21863v1
- Date: Wed, 26 Nov 2025 19:39:59 GMT
- Title: Saddle-Free Guidance: Improved On-Manifold Sampling without Labels or Additional Training
- Authors: Eric Yeats, Darryl Hannan, Wilson Fearn, Timothy Doster, Henry Kvinge, Scott Mahan,
- Abstract summary: We develop saddle-free guidance (SFG) which maintains estimates maximal positive curvature of the log density to guide individual score-based models.<n>Our experiments indicate that SFG achieves state-of-theart FID and FDDINOv2 metrics in single-model ImageNet512 generation.
- Score: 6.807078976578283
- License: http://creativecommons.org/licenses/by/4.0/
- Abstract: Score-based generative models require guidance in order to generate plausible, on-manifold samples. The most popular guidance method, Classifier-Free Guidance (CFG), is only applicable in settings with labeled data and requires training an additional unconditional score-based model. More recently, Auto-Guidance adopts a smaller, less capable version of the original model to guide generation. While each method effectively promotes the fidelity of generated data, each requires labeled data or the training of additional models, making it challenging to guide score-based models when (labeled) training data are not available or training new models is not feasible. We make the surprising discovery that the positive curvature of log density estimates in saddle regions provides strong guidance for score-based models. Motivated by this, we develop saddle-free guidance (SFG) which maintains estimates of maximal positive curvature of the log density to guide individual score-based models. SFG has the same computational cost of classifier-free guidance, does not require additional training, and works with off-the-shelf diffusion and flow matching models. Our experiments indicate that SFG achieves state-of-the-art FID and FD-DINOv2 metrics in single-model unconditional ImageNet-512 generation. When SFG is combined with Auto-Guidance, its unconditional samples achieve general state-of-the-art in FD-DINOv2 score. Our experiments with FLUX.1-dev and Stable Diffusion v3.5 indicate that SFG boosts the diversity of output images compared to CFG while maintaining excellent prompt adherence and image fidelity.
Related papers
- Diffusion Models without Classifier-free Guidance [41.59396565229466]
Model-guidance (MG) is a novel objective for training diffusion model addresses and removes commonly used guidance (CFG)<n>Our innovative approach transcends the standard modeling and incorporates the posterior probability of conditions.<n>Our method significantly accelerates the training process, doubles inference speed, and achieve exceptional quality that parallel surpass even concurrent diffusion models with CFG.
arXiv Detail & Related papers (2025-02-17T18:59:50Z) - Visual Generation Without Guidance [54.556375372833685]
We propose to build visual models that are free from sampling guided.<n>The resulting algorithm, Guidance-Free Training (GFT), matches the performance of CFG while reducing sampling to a single model, halving the cost.
arXiv Detail & Related papers (2025-01-26T06:48:05Z) - Gradient-Free Classifier Guidance for Diffusion Model Sampling [4.450496470631169]
Gradient-free Guidance (GFCG) method consistently improves class prediction accuracy.
For ImageNet 512$times$512, we achieve a record $FD_textDINOv2$ 23.09, while simultaneously attaining a higher classification Precision (94.3%) compared to ATG (90.2%)
arXiv Detail & Related papers (2024-11-23T00:22:21Z) - Guided Score identity Distillation for Data-Free One-Step Text-to-Image Generation [62.30570286073223]
Diffusion-based text-to-image generation models have demonstrated the ability to produce images aligned with textual descriptions.<n>We introduce a data-free guided distillation method that enables the efficient distillation of pretrained Diffusion models without access to the real training data.<n>By exclusively training with synthetic images generated by its one-step generator, our data-free distillation method rapidly improves FID and CLIP scores, achieving state-of-the-art FID performance while maintaining a competitive CLIP score.
arXiv Detail & Related papers (2024-06-03T17:44:11Z) - Exploring Beyond Logits: Hierarchical Dynamic Labeling Based on Embeddings for Semi-Supervised Classification [49.09505771145326]
We propose a Hierarchical Dynamic Labeling (HDL) algorithm that does not depend on model predictions and utilizes image embeddings to generate sample labels.
Our approach has the potential to change the paradigm of pseudo-label generation in semi-supervised learning.
arXiv Detail & Related papers (2024-04-26T06:00:27Z) - Discrete Diffusion Modeling by Estimating the Ratios of the Data Distribution [67.9215891673174]
We propose score entropy as a novel loss that naturally extends score matching to discrete spaces.
We test our Score Entropy Discrete Diffusion models on standard language modeling tasks.
arXiv Detail & Related papers (2023-10-25T17:59:12Z) - Consistency Regularization for Generalizable Source-free Domain
Adaptation [62.654883736925456]
Source-free domain adaptation (SFDA) aims to adapt a well-trained source model to an unlabelled target domain without accessing the source dataset.
Existing SFDA methods ONLY assess their adapted models on the target training set, neglecting the data from unseen but identically distributed testing sets.
We propose a consistency regularization framework to develop a more generalizable SFDA method.
arXiv Detail & Related papers (2023-08-03T07:45:53Z) - Consistent Diffusion Models: Mitigating Sampling Drift by Learning to be
Consistent [97.64313409741614]
We propose to enforce a emphconsistency property which states that predictions of the model on its own generated data are consistent across time.
We show that our novel training objective yields state-of-the-art results for conditional and unconditional generation in CIFAR-10 and baseline improvements in AFHQ and FFHQ.
arXiv Detail & Related papers (2023-02-17T18:45:04Z)
This list is automatically generated from the titles and abstracts of the papers in this site.
This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.