mlsanityData to 14 Oct 2026

Three Creates All: You Only Sample 3 Steps

arXiv 2603.22375 ↗First posted 23 Mar 2026cs.LG cs.AI cs.CVcited by 1

On the map: Diffusion models → Diffusion foundations and sampling

Abstract (arXiv)

Diffusion models deliver high-fidelity generation but remain slow at inference time due to many sequential network evaluations. We find that standard timestep conditioning becomes a key bottleneck for few-step sampling. Motivated by layer-dependent denoising dynamics, we propose Multi-layer Time Embedding Optimization (MTEO), which freeze the pretrained diffusion backbone and distill a small set of step-wise, layer-wise time embeddings from reference trajectories. MTEO is plug-and-play with existing ODE solvers, adds no inference-time overhead, and trains only a tiny fraction of parameters. Extensive experiments across diverse datasets and backbones show state-of-the-art performance in the few-step sampling and substantially narrow the gap between distillation-based and lightweight methods. Code will be available.

Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Placed in Diffusion foundations and sampling, one of 14 directions in Diffusion models. Builds on 39 of its 45 references that are in the map. First posted on arXiv 23 Mar 2026.

Builds on

39 of 45 references.

Closest ancestors · 26

Same direction, newest first

  1. 2025Distilling Parallel Gradients for Fast ODE Solvers of Diffusion Models20 Jul
  2. 2024Diffusion Sampling Correction via Approximately 10 Parameters10 Nov
  3. 2024Simple and Fast Distillation of Diffusion Models29 Sep
  4. 2024Consistency Models Made Easy20 Jun
  5. 2024Learning to Discretize Denoising Diffusion ODEs24 May
  6. 2024On the Trajectory Regularity of ODE-based Diffusion Sampling18 May
  7. 2024Align Your Steps: Optimizing Sampling Schedules in Diffusion Models22 Apr
  8. 2024Accelerating Diffusion Sampling with Optimized Time Steps27 Feb
  9. 2023Fast ODE-based Sampling for Diffusion Models in Around 5 Steps30 Nov
  10. 2023One-step Diffusion with Distribution Matching Distillation30 Nov
  11. 2023DPM-Solver-v3: Improved Diffusion ODE Solver with Empirical Model Statistics20 Oct
  12. 2023Consistency Trajectory Models: Learning Probability Flow ODE Trajectory of Diffusion1 Oct
14 more
  1. 2023AutoDiffusion: Training-Free Optimization of Time Steps and Architectures for Automated Diffusion Model Acceleration19 Sep
  2. 2023UniPC: A Unified Predictor-Corrector Framework for Fast Sampling of Diffusion Models9 Feb
  3. 2022Scalable Diffusion Models with Transformers19 Dec
  4. 2022DPM-Solver++: Fast Solver for Guided Sampling of Diffusion Probabilistic Models2 Nov
  5. 2022GENIE: Higher-Order Denoising Diffusion Solvers11 Oct
  6. 2022On Distillation of Guided Diffusion Models6 Oct
  7. 2022DPM-Solver: A Fast ODE Solver for Diffusion Probabilistic Model Sampling in Around 10 Steps2 Jun
  8. 2022Fast Sampling of Diffusion Models with Exponential Integrator29 Apr
  9. 2022Pseudo Numerical Methods for Diffusion Models on Manifolds20 Feb
  10. 2021High-Resolution Image Synthesis with Latent Diffusion Models20 Dec
  11. 2021Diffusion Models Beat GANs on Image Synthesis11 May
  12. 2020Score-Based Generative Modeling through Stochastic Differential Equations26 Nov
  13. 2020Denoising Diffusion Implicit Models6 Oct
  14. 2020Denoising Diffusion Probabilistic Models19 Jun

Other references · 13

  1. 2017Attention Is All You Needcited by 60,361
  2. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,744
  3. 2015U-Net: Convolutional Networks for Biomedical Image Segmentationcited by 19,713
  4. 2014Microsoft COCO: Common Objects in Contextcited by 17,608
  5. 2015Deep Unsupervised Learning using Nonequilibrium Thermodynamicscited by 7,058
  6. 2018A Style-Based Generator Architecture for Generative Adversarial Networkscited by 6,013
  7. 2019Generative Modeling by Estimating Gradients of the Data Distributioncited by 4,101
  8. 2016Improved Techniques for Training GANscited by 2,970
  9. 2021CLIPScore: A Reference-free Evaluation Metric for Image Captioningcited by 2,424
  10. 2015LSUN: Construction of a Large-scale Image Dataset using Deep Learning with Humans in the Loopcited by 1,355
  11. 2019Improved Precision and Recall Metric for Assessing Generative Modelscited by 944
  12. 2021Generating Images with Sparse Representationscited by 266
1 more
  1. 2022Transcormer: Transformer for Sentence Scoring with Sliding Language Modelingcited by 8

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 1.

1 paper.

  1. 2026Rethinking Prediction Horizons: Decoupling Model Queries from State Updates in Few-Step MeanFlowcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026On the Redundancy of Timestep Embeddings in Diffusion Models
  2. 2025Fewer Denoising Steps or Cheaper Per-Step Inference: Towards Compute-Optimal Diffusion Model Deployment
  3. 2026Leveraging Inference-Time Compute for Diffusion Models via Global Scheduling of Denoising Trajectories
  4. 2025DiffPro: Joint Timestep and Layer-Wise Precision Optimization for Efficient Diffusion Inference
  5. 2025You Only Look One Step: Accelerating Backpropagation in Diffusion Sampling with Gradient Shortcuts
  6. 2025Pretrained Diffusion Models Are Inherently Skipped-Step Samplers
  7. 2025Test-Time Scaling of Diffusion Models via Noise Trajectory Search
  8. 2026Timestep-Aware Block Masking for Efficient Diffusion Model Inference