Diffusion sampling from d-dimensional distributions requires at least ~sqrt(d) adaptive score queries when score estimates have polynomial accuracy.
hub Canonical reference
Consistency Models
Canonical reference. 75% of citing Pith papers cite this work as background.
48
Pith papers citing it
Background
75% of classified citations
abstract
Diffusion models have significantly advanced the fields of image, audio, and video generation, but they depend on an iterative sampling process that causes slow generation. To overcome this limitation, we propose consistency models, a new family of models that generate high quality samples by directly mapping noise to data. They support fast one-step generation by design, while still allowing multistep sampling to trade compute for sample quality. They also support zero-shot data editing, such as image inpainting, colorization, and super-resolution, without requiring explicit training on these tasks. Consistency models can be trained either by distilling pre-trained diffusion models, or as standalone generative models altogether. Through extensive experiments, we demonstrate that they outperform existing distillation techniques for diffusion models in one- and few-step sampling, achieving the new state-of-the-art FID of 3.55 on CIFAR-10 and 6.20 on ImageNet 64x64 for one-step generation. When trained in isolation, consistency models become a new family of generative models that can outperform existing one-step, non-adversarial generative models on standard benchmarks such as CIFAR-10, ImageNet 64x64 and LSUN 256x256.
hub tools
citation-role summary
background 6
method 2
citation-polarity summary
representative citing papers
Diffusion Policy models robot actions as a conditional diffusion process, outperforming prior state-of-the-art methods by 46.9% on average across 12 manipulation tasks from four benchmarks.
StreamingEffect enables real-time 720p human-centric video effect generation on one GPU via teacher-student distillation, keyframe control, and a new 130K video dataset.
ExpoCM enables fast one-step single-image HDR reconstruction via exposure-dependent perturbations and region-conditioned consistency trajectories derived from a probability flow ODE.
VLA models exhibit a compute-bound VLM phase followed by a memory-bound action phase on edge hardware; DP-Cache and V-AEFusion reduce redundancy and enable pipeline parallelism for up to 6x speedup on NPUs with marginal task degradation.
CoEdit is a zero-shot coopetitive framework for text-guided image editing that uses dual-entropy attention manipulation and entropic latent refinement to improve editing harmony and structural preservation.
Isokinetic Flow Matching adds a lightweight regularization term to flow matching that penalizes acceleration along paths via self-guided finite differences, yielding straighter trajectories and large gains in few-step sampling quality on CIFAR-10.
VOSR shows that competitive generative image super-resolution with faithful structures can be achieved by training a diffusion-style model from scratch on visual data alone, using a vision encoder for guidance and a restoration-oriented sampling strategy.
Early and late denoising steps in masked diffusion LMs are robust to smaller-model replacement, enabling 17% FLOPs reduction with modest generative quality loss.
Diff-ANO uses conditional consistency models and adjoint neural operator surrogates to enable fast, high-quality USCT reconstructions under sparse and partial views by replacing slow PDE solvers and enabling few-step sampling.
VideoRepair detects text-video misalignments via MLLM-generated questions and performs localized, region-preserving refinement to improve alignment in existing T2V diffusion models.
Shortcut models enable high-quality single or few-step sampling in diffusion models with one network and training phase by conditioning on desired step size.
Latent Consistency Models enable high-fidelity text-to-image generation in 2-4 steps by directly predicting solutions to the probability flow ODE in latent space, distilled from pre-trained LDMs.
CARV amortizes upstream diffusion teacher costs over noise resamples with timestep importance sampling and stratified-inverse-CDF sampling, delivering 2-3x effective compute gains in text-to-3D experiments and order-of-magnitude variance cuts in single-step distillation.
StreamGVE enables high-quality training-free video editing by converting the task to noise-to-data streaming generation with dual-branch fast sampling, self-attention bridges, cross-attention grounding, source-oriented guidance, and visual prompting.
REPA-P aligns intermediate representations in diffusion models with physical states using first-principles PDE residuals to accelerate convergence and boost out-of-distribution robustness on PDE tasks.
DCFold achieves AlphaFold3-level protein structure prediction accuracy in a single forward pass using Dual Consistency training and a Temporal Geodesic Matching scheduler, delivering 15x inference acceleration.
Pilot study uses pretrained video encoder features from lung ultrasound to predict 30-day CHF readmission, finding lower-lung views and temporal differences most informative with top MLP F1 of 0.80.
Decouples Sphere Encoder into fixed pretrained encoder and spherical latent denoiser, yielding higher quality and faster inference than the joint original on Animal-Faces, Oxford-Flowers and ImageNet-1K.
CLVR framework adds closed-loop visual verification, proxy prompt reinforcement learning, and delta-space weight merge to improve complex text-to-image generation over single-step or unverified multi-step baselines.
ZeNO frames noise optimization as a path-integral control problem solvable from zeroth-order reward evaluations, connecting to implicit Langevin dynamics for reward-tilted distributions.
dFlowGRPO is a new rate-aware RL method for discrete flow models that outperforms prior GRPO approaches on image generation and matches continuous flow models while supporting broad probability paths.
Tyche achieves competitive probabilistic weather forecasting skill and calibration using a single-step flow model with JVP-regularized training and rollout finetuning.
GCCM prevents shortcut collapse in consistency models for graph prediction by using contrastive negative pairs and input feature perturbation, leading to better performance than deterministic baselines.
citing papers explorer
-
Query Lower Bounds for Diffusion Sampling
Diffusion sampling from d-dimensional distributions requires at least ~sqrt(d) adaptive score queries when score estimates have polynomial accuracy.
-
Diffusion Policy: Visuomotor Policy Learning via Action Diffusion
Diffusion Policy models robot actions as a conditional diffusion process, outperforming prior state-of-the-art methods by 46.9% on average across 12 manipulation tasks from four benchmarks.
-
StreamingEffect: Real-Time Human-Centric Video Effect Generation
StreamingEffect enables real-time 720p human-centric video effect generation on one GPU via teacher-student distillation, keyframe control, and a new 130K video dataset.
-
ExpoCM: Exposure-Aware One-Step Generative Single-Image HDR Reconstruction
ExpoCM enables fast one-step single-image HDR reconstruction via exposure-dependent perturbations and region-conditioned consistency trajectories derived from a probability flow ODE.
-
Characterizing Vision-Language-Action Models across XPUs: Constraints and Acceleration for On-Robot Deployment
VLA models exhibit a compute-bound VLM phase followed by a memory-bound action phase on edge hardware; DP-Cache and V-AEFusion reduce redundancy and enable pipeline parallelism for up to 6x speedup on NPUs with marginal task degradation.
-
From Competition to Coopetition: Coopetitive Training-Free Image Editing Based on Text Guidance
CoEdit is a zero-shot coopetitive framework for text-guided image editing that uses dual-entropy attention manipulation and entropic latent refinement to improve editing harmony and structural preservation.
-
Isokinetic Flow Matching for Pathwise Straightening of Generative Flows
Isokinetic Flow Matching adds a lightweight regularization term to flow matching that penalizes acceleration along paths via self-guided finite differences, yielding straighter trajectories and large gains in few-step sampling quality on CIFAR-10.
-
VOSR: A Vision-Only Generative Model for Image Super-Resolution
VOSR shows that competitive generative image super-resolution with faithful structures can be achieved by training a diffusion-style model from scratch on visual data alone, using a vision encoder for guidance and a restoration-oriented sampling strategy.
-
Not All Denoising Steps Are Equal: Model Scheduling for Faster Masked Diffusion Language Models
Early and late denoising steps in masked diffusion LMs are robust to smaller-model replacement, enabling 17% FLOPs reduction with modest generative quality loss.
-
Diff-ANO: Towards Fast High-Resolution Ultrasound Computed Tomography via Conditional Consistency Models and Adjoint Neural Operators
Diff-ANO uses conditional consistency models and adjoint neural operator surrogates to enable fast, high-quality USCT reconstructions under sparse and partial views by replacing slow PDE solvers and enabling few-step sampling.
-
Self-Correcting Text-to-Video Generation with Misalignment Detection and Localized Refinement
VideoRepair detects text-video misalignments via MLLM-generated questions and performs localized, region-preserving refinement to improve alignment in existing T2V diffusion models.
-
One Step Diffusion via Shortcut Models
Shortcut models enable high-quality single or few-step sampling in diffusion models with one network and training phase by conditioning on desired step size.
-
Latent Consistency Models: Synthesizing High-Resolution Images with Few-Step Inference
Latent Consistency Models enable high-fidelity text-to-image generation in 2-4 steps by directly predicting solutions to the probability flow ODE in latent space, distilled from pre-trained LDMs.
-
Variance Reduction for Expectations with Diffusion Teachers
CARV amortizes upstream diffusion teacher costs over noise resamples with timestep importance sampling and stratified-inverse-CDF sampling, delivering 2-3x effective compute gains in text-to-3D experiments and order-of-magnitude variance cuts in single-step distillation.
-
StreamGVE: Training-Free Video Editing via Few-Step Streaming Video Generation
StreamGVE enables high-quality training-free video editing by converting the task to noise-to-data streaming generation with dual-branch fast sampling, self-attention bridges, cross-attention grounding, source-oriented guidance, and visual prompting.
-
Learning to Think in Physics: Breaking Shortcut Learning in Scientific Diffusion via Representation Alignment
REPA-P aligns intermediate representations in diffusion models with physical states using first-principles PDE residuals to accelerate convergence and boost out-of-distribution robustness on PDE tasks.
-
DCFold: Efficient Protein Structure Generation with Single Forward Pass
DCFold achieves AlphaFold3-level protein structure prediction accuracy in a single forward pass using Dual Consistency training and a Temporal Geodesic Matching scheduler, delivering 15x inference acceleration.
-
Prognostic Value of Lung Ultrasound Biomarkers for Readmission Risk in Congestive Heart Failure: A Pilot Data-Driven Analysis
Pilot study uses pretrained video encoder features from lung ultrasound to predict 30-day CHF readmission, finding lower-lung views and temporal differences most informative with top MLP F1 of 0.80.
-
Efficient Image Synthesis with Sphere Latent Encoder
Decouples Sphere Encoder into fixed pretrained encoder and spherical latent denoiser, yielding higher quality and faster inference than the joint original on Animal-Faces, Oxford-Flowers and ImageNet-1K.
-
Unlocking Complex Visual Generation via Closed-Loop Verified Reasoning
CLVR framework adds closed-loop visual verification, proxy prompt reinforcement learning, and delta-space weight merge to improve complex text-to-image generation over single-step or unverified multi-step baselines.
-
Gradient-Free Noise Optimization for Reward Alignment in Generative Models
ZeNO frames noise optimization as a path-integral control problem solvable from zeroth-order reward evaluations, connecting to implicit Langevin dynamics for reward-tilted distributions.
-
dFlowGRPO: Rate-Aware Policy Optimization for Discrete Flow Models
dFlowGRPO is a new rate-aware RL method for discrete flow models that outperforms prior GRPO approaches on image generation and matches continuous flow models while supporting broad probability paths.
-
Tyche: One Step Flow for Efficient Probabilistic Weather Forecasting
Tyche achieves competitive probabilistic weather forecasting skill and calibration using a single-step flow model with JVP-regularized training and rollout finetuning.
-
GCCM: Enhancing Generative Graph Prediction via Contrastive Consistency Model
GCCM prevents shortcut collapse in consistency models for graph prediction by using contrastive negative pairs and input feature perturbation, leading to better performance than deterministic baselines.
-
MetaSR: Content-Adaptive Metadata Orchestration for Generative Super-Resolution
MetaSR adaptively orchestrates metadata in a DiT-based generative SR model to deliver up to 1 dB PSNR gains and 50% bitrate savings across diverse content and degradations.
-
Pairing Regularization for Mitigating Many-to-One Collapse in GANs
Pairing regularization mitigates intra-mode collapse in GANs by penalizing redundant latent-to-sample mappings, improving recall under collapse-prone conditions or precision under stabilized training.
-
ELT: Elastic Looped Transformers for Visual Generation
Elastic Looped Transformers share weights across recurrent blocks and apply intra-loop self-distillation to deliver 4x parameter reduction while matching competitive FID and FVD scores on ImageNet and UCF-101.
-
A Unified View of Score-Based and Drifting Models
Drifting with Gaussian kernels exactly matches score-matching on smoothed distributions via Tweedie's formula, while Laplace kernels approximate this closely in high dimensions.
-
CoT-VLA: Visual Chain-of-Thought Reasoning for Vision-Language-Action Models
CoT-VLA is a 7B VLA that generates future visual frames autoregressively as planning goals before actions, outperforming prior VLAs by 17% on real-world tasks and 6% in simulation.
-
Unified Video Action Model
UVA learns a joint video-action latent representation with decoupled diffusion decoding heads, enabling a single model to perform accurate fast policy learning, forward/inverse dynamics, and video generation without performance loss versus task-specific methods.
-
Inference-Time Scaling for Diffusion Models beyond Scaling Denoising Steps
Diffusion models improve generation quality via inference-time search over noise candidates guided by verifiers and algorithms, yielding gains beyond denoising step scaling on class- and text-conditioned benchmarks.
-
DOLLAR: Few-Step Video Generation via Distillation and Latent Reward Optimization
DOLLAR combines variational score and consistency distillation for few-step video generation plus latent reward optimization, reporting 82.57 VBench score and up to 278x speedup over the teacher diffusion model for 128-frame 10-second videos.
-
Seed-TTS: A Family of High-Quality Versatile Speech Generation Models
Seed-TTS models produce speech matching human naturalness and speaker similarity, with added controllability via self-distillation and reinforcement learning.
-
Improved DDIM Sampling with Moment Matching Gaussian Mixtures
Moment-matched GMM kernels in DDIM yield lower FID and higher IS than Gaussian kernels at small sampling steps on CelebA-HQ, FFHQ, ImageNet, and Stable Diffusion tasks.
-
DiFaReli++: Diffusion Face Relighting with Consistent Cast Shadows
DiFaReli++ conditions a DDIM on shading references and inferred shadow maps to relight single-view faces with consistent shadows, trained only on 2D images and claiming SOTA on Multi-PIE.
-
CaloTrilogy: Toward a Breakthrough in One-Step, End-to-End, Physics-Guided Shower Generation for Modern Calorimeters
Presents CaloTrilogy, a unified one-step generative model for high-granularity calorimeter showers that combines velocity field integration, learned priors, and physics losses to match SOTA quality.
-
Lens: Rethinking Training Efficiency for Foundational Text-to-Image Models
Lens is a 3.8B-parameter text-to-image model that reaches competitive or superior performance to >6B-parameter systems using 19.3% of the training compute of Z-Image through a densely captioned 800M dataset, multi-resolution batching, semantic VAE, strong language encoder, RL fine-tuning, and 4-step
-
Teacher-Feature Drifting: One-Step Diffusion Distillation with Pretrained Diffusion Representations
A simplified one-step diffusion distillation uses pretrained teacher features directly for drifting loss plus a mode coverage term, achieving FID 1.58 on ImageNet-64 and 18.4 on SDXL.
-
Fast Text-to-Audio Generation with One-Step Sampling via Energy-Scoring and Auxiliary Contextual Representation Distillation
A one-step text-to-audio model using energy-distance training and contextual distillation outperforms prior fast baselines on AudioCaps and achieves up to 8.5x faster inference than the multi-step IMPACT system with competitive quality.
-
AdaCorrection: Adaptive Offset Cache Correction for Accurate Diffusion Transformers
AdaCorrection adaptively corrects offset caches in DiT inference via on-the-fly spatio-temporal validity checks to maintain near-original FID with moderate acceleration.
-
Towards Redundancy Reduction in Diffusion Models for Efficient Video Super-Resolution
OASIS reduces redundancy in diffusion models for real-world video super-resolution via attention specialization routing and progressive training, delivering state-of-the-art quality with 6.2x faster inference than prior one-step baselines.
-
A Unified Measure-Theoretic View of Diffusion, Score-Based, and Flow Matching Generative Models
Diffusion, score-based, and flow matching models are unified as instances of learning time-dependent vector fields inducing marginal distributions governed by continuity and Fokker-Planck equations.
-
OmniVLA-RL: A Vision-Language-Action Model with Spatial Understanding and Online RL
OmniVLA-RL uses a mix-of-transformers architecture and flow-matching reformulated as SDE with group segmented policy optimization to surpass prior VLA models on LIBERO benchmarks.
-
Discrete Meanflow Training Curriculum
A DMF curriculum initialized from pretrained flow models achieves one-step FID 3.36 on CIFAR-10 after only 2000 epochs by exploiting a discretized consistency property in the Meanflow objective.
-
Advancing Open-source World Models
LingBot-World is presented as an open-source world model that delivers high-fidelity simulation, minute-level contextual consistency, and real-time interactivity under one second latency.
-
World Models: A Comprehensive Survey of Architectures, Methodologies, Reasoning Paradigms, and Applications
The paper delivers a multi-axis taxonomy for world models that maps architectures, training families, reasoning strategies, and domains from early cognitive foundations through systems such as Dreamer, MuZero, and Sora while noting evaluation gaps.
- LIVEditor-14B: Lightning Unified Video Editing via In-Context Sparse Attention
- How to Guide Your Flow: Few-Step Alignment via Flow Map Reward Guidance