mlsanityData to 14 Oct 2026

Diffusion Lens: Interpreting Text Encoders in Text-to-Image Pipelines

arXiv 2403.05846 ↗First posted 9 Mar 2024cs.CV cs.CLcited by 30

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

Abstract (arXiv)

Text-to-image diffusion models (T2I) use a latent representation of a text prompt to guide the image generation process. However, the process by which the encoder produces the text representation is unknown. We propose the Diffusion Lens, a method for analyzing the text encoder of T2I models by generating images from its intermediate representations. Using the Diffusion Lens, we perform an extensive analysis of two recent T2I models. Exploring compound prompts, we find that complex scenes describing multiple objects are composed progressively and more slowly compared to simple scenes; Exploring knowledge retrieval, we find that representation of uncommon concepts requires further computation compared to common concepts, and that knowledge retrieval is gradual across layers. Overall, our findings provide valuable insights into the text encoder component in T2I pipelines.

Key facts. Cited by 30 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 29 of its 48 references that are in the map. First posted on arXiv 9 Mar 2024.

Builds on

29 of 48 references.

Closest ancestors · 5

Same direction, newest first

  1. 2023The Hidden Language of Diffusion Models1 Jun
  2. 2023Generating images of rare concepts using pre-trained diffusion models27 Apr
  3. 2022What the DAAM: Interpreting Stable Diffusion Using Cross Attention10 Oct
  4. 2022Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding23 May
  5. 2022Hierarchical Text-Conditional Image Generation with CLIP Latents13 Apr

Other references · 24

  1. 2017Attention Is All You Needcited by 60,361
  2. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,744
  3. 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,305
  4. 2021High-Resolution Image Synthesis with Latent Diffusion Modelscited by 18,863
  5. 2014Microsoft COCO: Common Objects in Contextcited by 17,608
  6. 2019Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformercited by 14,599
  7. 2022LAION-5B: An open large-scale dataset for training next generation image-text modelscited by 4,152
  8. 2022Locating and Editing Factual Associations in GPTcited by 2,691
  9. 2022Prompt-to-Prompt Image Editing with Cross Attention Controlcited by 2,431
  10. 2013Visualizing and Understanding Convolutional Networkscited by 2,290
  11. 2021Probing Classifiers: Promises, Shortcomings, and Advancescited by 876
  12. 2020A Primer in BERTology: What we know about how BERT workscited by 815
12 more
  1. 2020Stanza: A Python Natural Language Processing Toolkit for Many Human Languagescited by 723
  2. 2017Network Dissection: Quantifying Interpretability of Deep Visual Representationscited by 708
  3. 2023Eliciting Latent Predictions from Transformers with the Tuned Lenscited by 623
  4. 2022Transformer Feed-Forward Layers Build Predictions by Promoting Concepts in the Vocabulary Spacecited by 532
  5. 2023Towards Best Practices of Activation Patching in Language Models: Metrics and Methodscited by 359
  6. 2018Analysis Methods in Neural Language Processing: A Surveycited by 277
  7. 2023Interpreting CLIP's Image Representation via Text-Based Decompositioncited by 196
  8. 2021Post-hoc Interpretability for Neural NLP: A Surveycited by 156
  9. 2023Future Lens: Anticipating Subsequent Tokens from a Single Hidden Statecited by 121
  10. 2023Jump to Conclusions: Short-Cutting Transformers With Linear Transformationscited by 90
  11. 2023VISIT: Visualizing and Interpreting the Semantic Information Flow of Transformerscited by 32
  12. 2023ReFACT: Updating Text-to-Image Models by Editing the Text Encodercited by 32

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2024: 5; 2025: 13; 2026: 12.

30 papers.

Sort:

By area:

  1. 2024Golden Noise for Diffusion Models: A Learning Frameworkcited by 92
  2. 2025SAeUron: Interpretable Concept Unlearning in Diffusion Models with Sparse Autoencoderscited by 80
  3. 2026Verbalizable Representations Form a Global Workspace in Language Modelscited by 75
  4. 2024The Semantic Hub Hypothesis: Language Models Share Semantic Representations Across Languages and Modalitiescited by 69
  5. 2024Good Seed Makes a Good Crop: Discovering Secret Seeds in Text-to-Image Diffusion Modelscited by 48
  6. 2024textitRevelio: Interpreting and leveraging semantic information in diffusion modelscited by 30
  7. 2025PromptGuard: Soft Prompt-Guided Unsafe Content Moderation for Text-to-Image Modelscited by 17
  8. 2025TIDE : Temporal-Aware Sparse Autoencoders for Interpretable Diffusion Transformers in Image Generationcited by 16
  9. 2025When Are Concepts Erased From Diffusion Models?cited by 16
  10. 2025Beyond Transcription: Mechanistic Interpretability in ASRcited by 13
  11. 2025SAEmnesia: Erasing Concepts in Diffusion Models with Supervised Sparse Autoencoderscited by 10
  12. 2025Concept Steerers: Leveraging K-Sparse Autoencoders for Test-Time Controllable Generationscited by 7
18 more
  1. 2025VSC: Visual Search Compositional Text-to-Image Diffusion Modelcited by 7
  2. 2025Unboxing the Black Box: Mechanistic Interpretability for Algorithmic Understanding of Neural Networkscited by 6
  3. 2024NL-Eye: Abductive NLI for Imagescited by 5
  4. 2025Follow the Flow: On Information Flow Across Textual Tokens in Text-to-Image Modelscited by 4
  5. 2025Exploring Position Encoding in Diffusion U-Net for Training-free High-resolution Image Generationcited by 3
  6. 2026Localized Concept Erasure in Text-to-Image Diffusion Models via High-Level Representation Misdirectioncited by 3
  7. 2025Decoding Vision Transformers: the Diffusion Steering Lenscited by 2
  8. 2025BiasMap: Leveraging Cross-Attentions to Discover and Mitigate Hidden Social Biases in Text-to-Image Generationcited by 2
  9. 2026Interpretable Motion-Attentive Maps: Spatio-Temporally Localizing Concepts in Video Diffusion Transformerscited by 1
  10. 2026When Do Diffusion Models learn to Generate Multiple Objects?cited by 1
  11. 2026Vision-Language Binding in In-Context Image Generationcited by 1
  12. 2026Where Concept Erasure Should Occur: Concept-Layer Alignment in Text-to-Video Diffusion Modelscited by 1
  13. 2026Text-to-Image Models Need Less from Text Encoders Than You Thinkcited by 1
  14. 2026A Mechanistic Analysis of Adversarial Fine-tuning of Vision Transformerscited by 0
  15. 2026Intermediate Text Representation Guided Text-to-Image Generation for Enhancing One-and-Only Alignmentcited by 0
  16. 2026AEGIS: A Mechanism-Guided Defense against Visual Synonym Jailbreaks in Text-to-Image Modelscited by 0
  17. 2026Simile Understanding in Text-to-Image Models: An Evaluation Frameworkcited by 0
  18. 2026Learning to Read the Contextual Tokens in Diffusion Transformerscited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2023Prompt-Free Diffusion: Taking "Text" out of Text-to-Image Diffusion Models
  2. 2025Decoder-Only LLMs are Better Controllers for Diffusion Models
  3. 2023De-Diffusion Makes Text a Strong Cross-Modal Interface
  4. 2024DiffusionPID: Interpreting Diffusion via Partial Information Decomposition
  5. 2023CONFORM: Contrast is All You Need For High-Fidelity Text-to-Image Diffusion Models
  6. 2025CLIP is All You Need for Human-like Semantic Representations in Stable Diffusion
  7. 2023SUR-adapter: Enhancing Text-to-Image Pre-trained Diffusion Models with Large Language Models
  8. 2026Think-Then-Generate: Reasoning-Aware Text-to-Image Diffusion with LLM Encoders