mlsanityData to 14 Oct 2026

Improving Discriminator Guidance in Diffusion Models

arXiv 2503.16117 ↗First posted 20 Mar 2025cs.LGcited by 2

On the map: Diffusion models → Diffusion foundations and sampling

Abstract (arXiv)

Discriminator Guidance has become a popular method for efficiently refining pre-trained Score-Matching Diffusion models. However, in this paper, we demonstrate that the standard implementation of this technique does not necessarily lead to a distribution closer to the real data distribution. Specifically, we show that training the discriminator using Cross-Entropy loss, as commonly done, can in fact increase the Kullback-Leibler divergence between the model and target distributions, particularly when the discriminator overfits. To address this, we propose a theoretically sound training objective for discriminator guidance that properly minimizes the KL divergence. We analyze its properties and demonstrate empirically across multiple datasets that our proposed method consistently improves over the conventional method by producing samples of higher quality.

Key facts. Cited by 2 papers in the map (data through 14 Oct 2026). Placed in Diffusion foundations and sampling, one of 14 directions in Diffusion models. Builds on 24 of its 33 references that are in the map. First posted on arXiv 20 Mar 2025.

Builds on

24 of 33 references.

Closest ancestors · 11

Same direction, newest first

  1. 2023Mitigating Exposure Bias in Discriminator Guided Diffusion Models18 Nov
  2. 2023Discriminator Guidance for Autoregressive Diffusion Models24 Oct
  3. 2023Consistency Trajectory Models: Learning Probability Flow ODE Trajectory of Diffusion1 Oct
  4. 2023Restart Sampling for Improving Generative Processes26 Jun
  5. 2023DiffFit: Unlocking Transferability of Large Diffusion Models via Simple Parameter-Efficient Fine-Tuning13 Apr
  6. 2022Refining Generative Process with Discriminator Guidance in Score-based Diffusion Models28 Nov
  7. 2022Elucidating the Design Space of Diffusion-Based Generative Models1 Jun
  8. 2021Gotta Go Fast When Generating Data with Score-Based Models28 May
  9. 2021Diffusion Models Beat GANs on Image Synthesis11 May
  10. 2021Maximum Likelihood Training of Score-Based Diffusion Models22 Jan
  11. 2020Score-Based Generative Modeling through Stochastic Differential Equations26 Nov

Other references · 13

  1. 2015U-Net: Convolutional Networks for Biomedical Image Segmentationcited by 19,713
  2. 2022Imagen Video: High Definition Video Generation with Diffusion Modelscited by 1,778
  3. 2020DiffWave: A Versatile Diffusion Model for Audio Synthesiscited by 1,354
  4. 2019Improved Precision and Recall Metric for Assessing Generative Modelscited by 944
  5. 2016f-GAN: Training Generative Neural Samplers using Variational Divergence Minimizationcited by 506
  6. 2019Sliced Score Matching: A Scalable Approach to Density and Score Estimationcited by 384
  7. 2023Generative Diffusion Models on Graphs: Methods and Applicationscited by 81
  8. 2020Your GAN is Secretly an Energy-based Model and You Should use Discriminator Driven Latent Samplingcited by 71
  9. 2018Discriminator Rejection Samplingcited by 49
  10. 2020Refining Deep Generative Models via Discriminator Gradient Flowcited by 45
  11. 2016Generative Adversarial Nets from a Density Ratio Estimation Perspectivecited by 24
  12. 2023Precision-Recall Divergence Optimization for Generative Modeling with GANs and Normalizing Flowscited by 9
1 more
  1. 2023Optimal Budgeted Rejection Sampling for Generative Modelscited by 5

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 2.

2 papers.

  1. 2026Learning To Sample From Diffusion Models Via Inverse Reinforcement Learningcited by 0
  2. 2026Adversarial Learning of Classifier-Free Guidance Schedulescited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026Provably Reliable Classifier Guidance via Cross-Entropy Control
  2. 2023Bridging the Gap: Addressing Discrepancies in Diffusion Model Training for Classifier-Free Guidance
  3. 2026Distilling Diffusion Score Discrepancy for Efficient Training Data Attribution
  4. 2025Guiding Noisy Label Conditional Diffusion Models with Score-based Discriminator Correction
  5. 2026Tightening the Score Matching Gap for Diffusion Models
  6. 2024Theoretical Insights for Diffusion Guidance: A Case Study for Gaussian Mixture Models
  7. 2026Diffusion Models Observe Only Gradients: A Geometric Perspective on Score Matching Errors
  8. 2025Towards a unified framework for guided diffusion models