mlsanityData to 14 Oct 2026

Plug-and-Play Guidance for Discrete Diffusion Models via Gradient-Informed Logit Correction

arXiv 2606.06303 ↗First posted 4 Jun 2026cs.LG cs.AIcited by 3

On the map: Diffusion models → Diffusion foundations and sampling

Abstract (arXiv)

Controllable generation with discrete diffusion models is often hindered by high computational overhead or the need for retraining. In this paper, we present underlineGradient-underlineInformed underlineLogit underlineCorrection (GILC), a plug-and-play framework that efficiently estimates guidance signals by repurposing the pretrained denoising network as a variational proxy. To circumvent the gradient instability inherent in high-dimensional discrete spaces, we introduce a Jacobian-free mechanism that directly corrects the clean prediction logits, facilitating stable and effective guidance. Our method accommodates both differentiable and non-differentiable reward functions. Extensive experiments across DNA, protein sequence, and molecular generation tasks demonstrate that GILC achieves state-of-the-art performance without additional training, frequently outperforming fine-tuning approaches.

Key facts. Cited by 3 papers in the map (data through 14 Oct 2026). Placed in Diffusion foundations and sampling, one of 14 directions in Diffusion models. Builds on 50 of its 71 references that are in the map. First posted on arXiv 4 Jun 2026.

Builds on

50 of 71 references.

Closest ancestors · 15

Same direction, newest first

  1. 2025Inference-Time Scaling of Discrete Diffusion Models via Importance Weighting and Optimal Proposal Design28 May
  2. 2025Variational Autoencoding Discrete Diffusion with Enhanced Dimensional Correlations Modeling23 May
  3. 2024Simple Guidance Mechanisms for Discrete Diffusion Models13 Dec
  4. 2024Derivative-Free Guidance in Continuous and Discrete Diffusion Models with Soft Value-Based Decoding15 Aug
  5. 2024Amortizing intractable inference in diffusion models for vision, language, and control31 May
  6. 2024Scaling Rectified Flow Transformers for High-Resolution Image Synthesis5 Mar
  7. 2023Manifold Preserving Guided Diffusion28 Nov
  8. 2023Universal Guidance for Diffusion Models14 Feb
  9. 2022Classifier-Free Diffusion Guidance26 Jul
  10. 2022A Continuous Time Framework for Discrete Denoising Models30 May
  11. 2021High-Resolution Image Synthesis with Latent Diffusion Models20 Dec
  12. 2021Structured Denoising Diffusion Models in Discrete State-Spaces7 Jul
3 more
  1. 2021Diffusion Models Beat GANs on Image Synthesis11 May
  2. 2020Score-Based Generative Modeling through Stochastic Differential Equations26 Nov
  3. 2020Denoising Diffusion Probabilistic Models19 Jun

Other references · 35

  1. 2024DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Modelscited by 9,379
  2. 2023Direct Preference Optimization: Your Language Model is Secretly a Reward Modelcited by 8,065
  3. 2022DreamFusion: Text-to-3D using 2D Diffusioncited by 3,025
  4. 2016Categorical Reparameterization with Gumbel-Softmaxcited by 2,739
  5. 2013Estimating or Propagating Gradients Through Stochastic Neurons for Conditional Computationcited by 2,079
  6. 2022Diffusion Posterior Sampling for General Noisy Inverse Problemscited by 1,416
  7. 2023Training Diffusion Models with Reinforcement Learningcited by 982
  8. 2021E(n) Equivariant Graph Neural Networkscited by 965
  9. 2025Large Language Diffusion Modelscited by 912
  10. 2024Simple and Effective Masked Diffusion Language Modelscited by 848
  11. 2019Advantage-Weighted Regression: Simple and Scalable Off-Policy Reinforcement Learningcited by 782
  12. 2022Equivariant Diffusion for Molecule Generation in 3Dcited by 677
23 more
  1. 2024Simplified and Generalized Masked Diffusion for Discrete Datacited by 626
  2. 2022DiGress: Discrete Denoising diffusion for graph generationcited by 520
  3. 2024Sora: A Review on Background, Technology, Limitations, and Opportunities of Large Vision Modelscited by 477
  4. 2023Directly Fine-Tuning Diffusion Models on Differentiable Rewardscited by 448
  5. 2024Generative Flows on Discrete State-Spaces: Enabling Multimodal Flows with Applications to Protein Co-Designcited by 305
  6. 2023Aligning Text-to-Image Diffusion Models with Reward Backpropagationcited by 288
  7. 2023Practical and Asymptotically Exact Conditional Sampling in Diffusion Modelscited by 232
  8. 2024Unlocking Guidance for Discrete State-Space Diffusion and Flow Modelscited by 129
  9. 2024Theoretical guarantees on the best-of-n alignment policycited by 124
  10. 2023Delta Denoising Scorecited by 121
  11. 2024Understanding Reinforcement Learning-Based Fine-Tuning of Diffusion Models: A Tutorial and Reviewcited by 98
  12. 2021Inverse design of 3d molecular structures with conditional generative neural networkscited by 93
  13. 2021Oops I Took A Gradient: Scalable Sampling for Discrete Distributionscited by 92
  14. 2024Fine-Tuning Discrete Diffusion Models via Reward Optimization with Applications to DNA and Protein Designcited by 79
  15. 2022Equivariant Energy-Guided SDE for Inverse Molecular Designcited by 73
  16. 2021Estimating High Order Gradients of the Data Distribution by Denoisingcited by 64
  17. 2025Fine-Tuning Discrete Diffusion Models with Policy Gradient Methodscited by 38
  18. 2024Steering Masked Discrete Diffusion Models via Discrete Denoising Posterior Predictioncited by 32
  19. 2022Bidirectional Generation of Structure and Properties Through a Single Molecular Foundation Modelcited by 25
  20. 2025Split Gibbs Discrete Diffusion Posterior Samplingcited by 18
  21. 2025Constrained Discrete Diffusioncited by 17
  22. 2015Linear Response Methods for Accurate Covariance Estimates from Mean Field Variational Bayescited by 14
  23. 2025TFG-Flow: Training-free Guidance in Multimodal Generative Flowcited by 5

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 3.

3 papers.

  1. 2026VGAS: Variance-Reduced Guidance and Adaptive Selection for Training-Free Reward Alignment in Discrete Diffusioncited by 1
  2. 2026ProtoGuide: Prototype-Driven Guidance for Class-Conditional Graph Generationcited by 0
  3. 2026Grab a Coffee: Future-Aware Guidance for Discrete Diffusion with Compiled Objectivescited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025Tortoise and Hare Guidance: Accelerating Diffusion Model Inference with Multirate Integration
  2. 2025Token Perturbation Guidance for Diffusion Models
  3. 2025Training-Free Guidance Beyond Differentiability: Scalable Path Steering with Tree Search in Diffusion and Flow Models
  4. 2026Spectral Guidance for Flexible and Efficient Control of Diffusion Models
  5. 2026Value-Informed Schedule Trimming for Accelerated Sequential Monte Carlo Guidance in Discrete Diffusion
  6. 2026IDRF: Inverse-Distilled Reward Fine-tuning of Masked Discrete Diffusion Models
  7. 2026On the Robustness of Distribution Support under Diffusion Guidance
  8. 2026Error as Signal: Stiffness-Aware Diffusion Sampling via Embedded Runge-Kutta Guidance