REVIEW 6 cited by
Stable Consistency Tuning: Understanding and Improving Consistency Models
Not yet reviewed by Pith; the record is open.
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
Signed reviews
read the original abstract
Diffusion models achieve superior generation quality but suffer from slow generation speed due to the iterative nature of denoising. In contrast, consistency models, a new generative family, achieve competitive performance with significantly faster sampling. These models are trained either through consistency distillation, which leverages pretrained diffusion models, or consistency training/tuning directly from raw data. In this work, we propose a novel framework for understanding consistency models by modeling the denoising process of the diffusion model as a Markov Decision Process (MDP) and framing consistency model training as the value estimation through Temporal Difference~(TD) Learning. More importantly, this framework allows us to analyze the limitations of current consistency training/tuning strategies. Built upon Easy Consistency Tuning (ECT), we propose Stable Consistency Tuning (SCT), which incorporates variance-reduced learning using the score identity. SCT leads to significant performance improvements on benchmarks such as CIFAR-10 and ImageNet-64. On ImageNet-64, SCT achieves 1-step FID 2.42 and 2-step FID 1.55, a new SoTA for consistency models.
Forward citations
Cited by 6 Pith papers
-
Sample-Adaptive Latent Rewards for Uncertainty-Guided Diffusion Post-Training
SURE learns sample-adaptive variance in a latent reward model and uses that variance to weight dense post-training feedback, improving image and video diffusion alignment in reported experiments.
-
Amortized Moment Matching for Visual Generation
Amortized Fréchet Distance uses neural nets to match conditional means and covariances, yielding stronger one-step visual generators than explicit FD-loss or multi-step teachers.
-
Understanding, Accelerating, and Improving MeanFlow Training
Training MeanFlow by first forming instantaneous velocity and short-gap average velocity, then shifting to long gaps, improves 1-NFE ImageNet FID from 3.43 to 2.87 and speeds training by about 2.5x.
-
Transition Models: Rethinking the Generative Learning Objective
TiM trains a single diffusion-type model on arbitrary time-interval transitions, achieving strong one-step and multi-step text-to-image generation with 865M parameters.
-
Modular MeanFlow: Towards Stable and Scalable One-Step Generative Modeling
Modular MeanFlow interpolates between stop-gradient and full-gradient training for one-step generative models via a warmup schedule, reporting a CIFAR-10 FID of 3.41.
-
Align Your Flow: Scaling Continuous-Time Flow Map Distillation
Align Your Flow distills flow maps with new continuous-time objectives and autoguidance, achieving state-of-the-art few-step FID on ImageNet and strong text-to-image results.
Discussion (0). Continue with ORCID to comment.