mlsanityData to 7 Oct 2026

InfoDiffusion: Representation Learning Using Information Maximizing Diffusion Models

arXiv 2306.08757 ↗First posted 14 Jun 2023cs.LG cs.CVcited by 48

On the map: Diffusion models → Diffusion foundations and sampling

Abstract (arXiv)

While diffusion models excel at generating high-quality samples, their latent variables typically lack semantic meaning and are not suitable for representation learning. Here, we propose InfoDiffusion, an algorithm that augments diffusion models with low-dimensional latent variables that capture high-level factors of variation in the data. InfoDiffusion relies on a learning objective regularized with the mutual information between observed and hidden variables, which improves latent space quality and prevents the latents from being ignored by expressive diffusion-based decoders. Empirically, we find that InfoDiffusion learns disentangled and human-interpretable latent representations that are competitive with state-of-the-art generative and contrastive methods, while retaining the high sample quality of diffusion models. Our method enables manipulating the attributes of generated images and has the potential to assist tasks that require exploring a learned latent space to generate quality samples, e.g., generative design.

Key facts. Cited by 48 papers in the map (data through 7 Oct 2026). Placed in Diffusion foundations and sampling, one of 14 directions in Diffusion models. Builds on 26 of its 36 references that are in the map. First posted on arXiv 14 Jun 2023.

Builds on

26 of 36 references.

Closest ancestors · 6

Same direction, newest first

  1. 2022Unsupervised Representation Learning from Pre-trained Diffusion Probabilistic Models26 Dec
  2. 2022Diffusion Models: A Comprehensive Survey of Methods and Applications2 Sep
  3. 2021High-Resolution Image Synthesis with Latent Diffusion Models20 Dec
  4. 2021Diffusion Autoencoders: Toward a Meaningful and Decodable Representation30 Nov
  5. 2021Diffusion Models Beat GANs on Image Synthesis11 May
  6. 2020Denoising Diffusion Probabilistic Models19 Jun

Other references · 20

  1. 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,160
  2. 2015U-Net: Convolutional Networks for Biomedical Image Segmentationcited by 19,604
  3. 2020A Simple Framework for Contrastive Learning of Visual Representationscited by 11,979
  4. 2021Emerging Properties in Self-Supervised Vision Transformerscited by 6,893
  5. 2022Hierarchical Text-Conditional Image Generation with CLIP Latentscited by 6,539
  6. 2022Photorealistic Text-to-Image Diffusion Models with Deep Language Understandingcited by 6,322
  7. 2018A Style-Based Generator Architecture for Generative Adversarial Networkscited by 5,986
  8. 2022Generative Adversarial Networkscited by 5,603
  9. 2017Fashion-MNIST: a Novel Image Dataset for Benchmarking Machine Learning Algorithmscited by 4,283
  10. 2014Deep Learning Face Attributes in the Wildcited by 3,679
  11. 2016Categorical Reparameterization with Gumbel-Softmaxcited by 2,708
  12. 2020Improved Baselines with Momentum Contrastive Learningcited by 2,282
8 more
  1. 2020DiffWave: A Versatile Diffusion Model for Audio Synthesiscited by 1,339
  2. 2018Challenging Common Assumptions in the Unsupervised Learning of Disentangled Representationscited by 929
  3. 2016InfoGAN: Interpretable Representation Learning by Information Maximizing Generative Adversarial Netscited by 858
  4. 2022GeoDiff: a Geometric Diffusion Model for Molecular Conformation Generationcited by 511
  5. 2015Adversarial Autoencoderscited by 359
  6. 2022Torsional Diffusion for Molecular Conformer Generationcited by 262
  7. 2017InfoVAE: Information Maximizing Variational Autoencoderscited by 124
  8. 2022NashAE: Disentangling Representations through Adversarial Covariance Minimizationcited by 13

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2023: 5; 2024: 19; 2025: 16; 2026: 8.

48 papers.

Sort:

By area:

  1. 2024Simple and Effective Masked Diffusion Language Modelscited by 800
  2. 2025Block Diffusion: Interpolating Between Autoregressive and Diffusion Language Modelscited by 385
  3. 2024Simple Guidance Mechanisms for Discrete Diffusion Modelscited by 148
  4. 2023SODA: Bottleneck Diffusion Models for Representation Learningcited by 76
  5. 2023Self-Discovering Interpretable Diffusion Latent Directions for Responsible Text-to-Image Generationcited by 69
  6. 2024Diffusion Models and Representation Learning: A Surveycited by 53
  7. 2023Diffusion Models With Learned Adaptive Noisecited by 50
  8. 2024Diffusion Model with Cross Attention as an Inductive Bias for Disentanglementcited by 27
  9. 2024Training Unbiased Diffusion Models From Biased Datasetcited by 25
  10. 2025Encoder-Decoder Diffusion Language Models for Efficient Training and Inferencecited by 21
  11. 2023IMPUS: Image Morphing with Perceptually-Uniform Sampling Using Diffusion Modelscited by 20
  12. 2024DisCo-Diff: Enhancing Continuous Diffusion Models with Discrete Latentscited by 20
36 more
  1. 2025Understanding Representation Dynamics of Diffusion Models via Low-Dimensional Modelingcited by 18
  2. 2026Learn from Your Mistakes: Self-Correcting Masked Diffusion Modelscited by 12
  3. 2024Regularized Conditional Diffusion Model for Multi-Task Preference Alignmentcited by 9
  4. 2026DUEL: Exact Likelihood for Masked Diffusion via Deterministic Unmaskingcited by 8
  5. 2024Diffusion Bridge AutoEncoders for Unsupervised Representation Learningcited by 6
  6. 2025Diffusion Generative Modeling on Lie Group Representationscited by 6
  7. 2024Unsupervised Composable Representations for Audiocited by 4
  8. 2024Disentangling Disentangled Representations: Towards Improved Latent Units via Diffusion Modelscited by 4
  9. 2025Vision-Language-Vision Auto-Encoder: Scalable Knowledge Distillation from Diffusion Modelscited by 4
  10. 2026Noise Scheduling as Information-Guided Allocation in Diffusion Trainingcited by 4
  11. 2025FedDifRC: Unlocking the Potential of Text-to-Image Diffusion Models in Heterogeneous Federated Learningcited by 3
  12. 2025Denoising Multi-Beta VAE: Representation Learning for Disentanglement and Generationcited by 3
  13. 2024Denoising Diffusion Variational Inference: Diffusion Models as Expressive Variational Posteriorscited by 2
  14. 2024ParamReL: Learning Parameter Space Representation via Progressively Encoding Bayesian Flow Networkscited by 2
  15. 2024Multimodal ELBO with Diffusion Decoderscited by 2
  16. 2024Channel-aware Contrastive Conditional Diffusion for Multivariate Probabilistic Time Series Forecastingcited by 2
  17. 2024Unsupervised Representation Learning from Sparse Transformation Analysiscited by 2
  18. 2024Feature-guided score diffusion for sampling conditional densitiescited by 2
  19. 2024TreeDiffusion: Hierarchical Generative Clustering for Conditional Diffusioncited by 2
  20. 2026Set Diffusion: Interpolating Token Orderings Between Autoregression and Diffusion for Fast and Flexible Decodingcited by 2
  21. 2024Improving Training-free Conditional Diffusion Model via Fisher Informationcited by 1
  22. 2024Diffusion-Based Generation of Neural Activity from Disentangled Latent Codescited by 1
  23. 2025Graffe: Graph Representation Learning via Diffusion Probabilistic Modelscited by 1
  24. 2025Addressing degeneracies in latent interpolation for diffusion modelscited by 1
  25. 2025On Designing Diffusion Autoencoders for Efficient Generation and Representation Learningcited by 1
  26. 2025Compositional Discrete Latent Code for High Fidelity, Productive Diffusion Modelscited by 1
  27. 2025Cross-modal Causal Intervention for Alzheimer's Disease Predictioncited by 1
  28. 2025Coupled Flow Matchingcited by 1
  29. 2026Evaluating the Representation Space of Diffusion Models via Self-Supervised Principlescited by 1
  30. 2023Diffusion Based Causal Representation Learningcited by 0
  31. 2025Patronus: Interpretable Diffusion Models with Prototypescited by 0
  32. 2025Revisiting Diffusion Autoencoder Training for Image Reconstruction Qualitycited by 0
  33. 2025Disentangled representations via score-based variational autoencoderscited by 0
  34. 2026Semantic Generative Tuning for Unified Multimodal Modelscited by 0
  35. 2026Learning a Maximum Entropy Model for Visual Textures using Diffusioncited by 0
  36. 2026CenSynCMB: Centre Maps and Physics-Guided Synthesis for Microbleed Detectioncited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2024Nested Diffusion Models Using Hierarchical Latent Priors
  2. 2026Diffusing in the Right Space: A Systematic Study of Latent Diffusability
  3. 2025Continuously Augmented Discrete Diffusion model for Categorical Generative Modeling
  4. 2024Decoding Diffusion: A Scalable Framework for Unsupervised Analysis of Latent Space Biases and Representations Using Natural Language Prompts
  5. 2024Diffusion Models Need Visual Priors for Image Generation
  6. 2023Interpretable Diffusion via Information Decomposition
  7. 2025Automated Learning of Semantic Embedding Representations for Diffusion Models
  8. 2022Let us Build Bridges: Understanding and Extending Diffusion Generative Models