mlsanityData to 14 Oct 2026

Concept-TRAK: Understanding how diffusion models learn concepts through concept-level attribution

arXiv 2507.06547 ↗First posted 9 Jul 2025cs.CV cs.LGcited by 3

On the map: Diffusion models → Image generation, text-to-image and editing

Abstract (arXiv)

While diffusion models excel at image generation, their growing adoption raises critical concerns about copyright issues and model transparency. Existing attribution methods identify training examples influencing an entire image, but fall short in isolating contributions to specific elements, such as styles or objects, that are of primary concern to stakeholders. To address this gap, we introduce concept-level attribution through a novel method called Concept-TRAK, which extends influence functions with a key innovation: specialized training and utility loss functions designed to isolate concept-specific influences rather than overall reconstruction quality. We evaluate Concept-TRAK on novel concept attribution benchmarks using Synthetic and CelebA-HQ datasets, as well as the established AbC benchmark, showing substantial improvements over prior methods in concept-level attribution scenarios. We further demonstrate its versatility on real-world text-to-image generation with compositional and multi-concept prompts.

Key facts. Cited by 3 papers in the map (data through 14 Oct 2026). Placed in Image generation, text-to-image and editing, one of 14 directions in Diffusion models. Builds on 58 of its 65 references that are in the map. First posted on arXiv 9 Jul 2025.

Builds on

58 of 65 references.

Closest ancestors · 7

Same direction, newest first

  1. 2023Concept Sliders: LoRA Adaptors for Precise Control in Diffusion Models20 Nov
  2. 2023Concept Algebra for (Score-Based) Text-Controlled Generative Models7 Feb
  3. 2022Diffusion Art or Digital Forgery? Investigating Data Replication in Diffusion Models7 Dec
  4. 2022Compositional Visual Generation with Composable Diffusion Models3 Jun
  5. 2022Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding23 May
  6. 2022Hierarchical Text-Conditional Image Generation with CLIP Latents13 Apr
  7. 2021GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Models20 Dec

Other references · 51

  1. 2020Denoising Diffusion Probabilistic Modelscited by 21,076
  2. 2021High-Resolution Image Synthesis with Latent Diffusion Modelscited by 18,863
  3. 2020Denoising Diffusion Implicit Modelscited by 9,781
  4. 2020Score-Based Generative Modeling through Stochastic Differential Equationscited by 9,122
  5. 2025Qwen3 Technical Reportcited by 8,820
  6. 2023Direct Preference Optimization: Your Language Model is Secretly a Reward Modelcited by 8,065
  7. 2015Deep Unsupervised Learning using Nonequilibrium Thermodynamicscited by 7,058
  8. 2022Classifier-Free Diffusion Guidancecited by 5,847
  9. 2023SDXL: Improving Latent Diffusion Models for High-Resolution Image Synthesiscited by 4,356
  10. 2024Scaling Rectified Flow Transformers for High-Resolution Image Synthesiscited by 4,181
  11. 2022LAION-5B: An open large-scale dataset for training next generation image-text modelscited by 4,152
  12. 2022An Image is Worth One Word: Personalizing Text-to-Image Generation using Textual Inversioncited by 2,245
38 more
  1. 2017Understanding Black-box Predictions via Influence Functionscited by 1,741
  2. 2022Diffusion Posterior Sampling for General Noisy Inverse Problemscited by 1,416
  3. 2023Diffusion Model Alignment Using Direct Preference Optimizationcited by 790
  4. 2023Extracting Training Data from Diffusion Modelscited by 732
  5. 2022Pseudo Numerical Methods for Diffusion Models on Manifoldscited by 700
  6. 2019Data Shapley: Equitable Valuation of Data for Machine Learningcited by 629
  7. 2020Estimating Training Data Influence by Tracing Gradient Descentcited by 580
  8. 2022Improving Diffusion Models for Inverse Problems using Manifold Constraintscited by 508
  9. 2021Flow Network based Generative Models for Non-Iterative Diverse Candidate Generationcited by 392
  10. 2023Studying Large Language Model Generalization with Influence Functionscited by 326
  11. 2021GFlowNet Foundationscited by 290
  12. 2023TRAK: Attributing Model Behavior at Scalecited by 283
  13. 2019Towards Efficient Data Valuation Based on the Shapley Valuecited by 281
  14. 2023Understanding and Mitigating Copying in Diffusion Modelscited by 209
  15. 2021A Variational Perspective on Diffusion-Based Generative Models and Score Matchingcited by 183
  16. 2024CFG++: Manifold-constrained Classifier Free Guidance for Diffusion Modelscited by 149
  17. 2019Efficient Task-Specific Data Valuation for Nearest Neighbor Algorithmscited by 148
  18. 2023Understanding the Latent Space of Diffusion Models through the Lens of Riemannian Geometrycited by 125
  19. 2016Sequence Tutor: Conservative Fine-Tuning of Sequence Generation Models with KL-controlcited by 101
  20. 2024Detecting, Explaining, and Mitigating Memorization in Diffusion Modelscited by 95
  21. 2024What is Your Data Worth to GPT? LLM-Scale Data Valuation with Influence Functionscited by 94
  22. 2023Compositional Abilities Emerge Multiplicatively: Exploring Diffusion Models on a Synthetic Taskcited by 93
  23. 2022On Reinforcement Learning and Distribution Matching for Fine-Tuning Language Models with no Catastrophic Forgettingcited by 82
  24. 2024Data Shapley in One Training Runcited by 81
  25. 2024FasterDiT: Towards Faster Diffusion Transformers Training without Architecture Modificationcited by 64
  26. 2024No Training, No Problem: Rethinking Classifier-Free Guidance for Diffusion Modelscited by 53
  27. 2023Intriguing Properties of Data Attribution on Diffusion Modelscited by 47
  28. 2017Assigning a value to a power likelihood in a general Bayesian modelcited by 46
  29. 2023Evaluating Data Attribution for Text-to-Image Modelscited by 45
  30. 2024Procedural Knowledge in Pretraining Drives Reasoning in Large Language Modelscited by 41
  31. 2024Influence Functions for Scalable Data Attribution in Diffusion Modelscited by 33
  32. 2024Most Influential Subset Selection: Challenges, Promises, and Beyondcited by 32
  33. 2024Efficient Diversity-Preserving Diffusion Alignment via Gradient-Informed GFlowNetscited by 32
  34. 2024Data Attribution for Text-to-Image Models by Unlearning Synthesized Imagescited by 22
  35. 2024Diffusion Attribution Score: Evaluating Training Data Influence in Diffusion Modelscited by 22
  36. 2023Threshold KNN-Shapley: A Linear-Time and Privacy-Friendly Approach to Data Valuationcited by 19
  37. 2025TCFG: Tangential Damping Classifier-free Guidancecited by 15
  38. 2024Data Attribution for Diffusion Models: Timestep-induced Bias in Influence Estimationcited by 14

50 most cited of 51 All 51 (JSON)

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 3.

3 papers.

  1. 2026Motion Attribution for Video Generationcited by 5
  2. 2026TrueMuse: A Benchmark for Data Attribution in Text-to-Music Modelscited by 0
  3. 2026Attributing HOW, Not Just WHICH: Counterfactual Response Trajectories for Diffusion Modelscited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2024Unveiling Concept Attribution in Diffusion Models
  2. 2023Ablating Concepts in Text-to-Image Diffusion Models
  3. 2023The Hidden Language of Diffusion Models
  4. 2025Concept Reachability in Diffusion Models: Beyond Dataset Constraints
  5. 2024Holistic Unlearning Benchmark: A Multi-Faceted Evaluation for Text-to-Image Diffusion Model Unlearning
  6. 2025Blending Concepts with Text-to-Image Diffusion Models
  7. 2024Detecting Origin Attribution for Text-to-Image Diffusion Models
  8. 2023On Copyright Risks of Text-to-Image Diffusion Models