mlsanityData to 14 Oct 2026

Geometrical Properties of Text Token Embeddings for Strong Semantic Binding in Text-to-Image Generation

arXiv 2503.23011 ↗First posted 29 Mar 2025cs.CV cs.AIcited by 5

Not on the map. Why?

Abstract (arXiv)

Text-to-image (T2I) models often suffer from text-image misalignment in complex scenes involving multiple objects and attributes. Semantic binding has attempted to associate the generated attributes and objects with their corresponding noun phrases (NPs) by text or latent optimizations with the modulation of cross-attention (CA) maps; yet, the factors that influence semantic binding remain underexplored. Here, we investigate the geometrical properties of text token embeddings and their CA maps. We found that the geometrical properties of token embeddings, specifically angular distances and norms, are crucial factors in the differentiation of the CA map. These theoretical findings led to our proposed training-free text-embedding-aware T2I framework, dubbed TokeBi, for strong semantic binding. TokeBi consists of Causality-Aware Projection-Out (CAPO) for distinguishing inter-NP CA maps and Adaptive Token Mixing (ATM) for enhancing inter-NP separation while maintaining intra-NP cohesion in CA maps. Extensive experiments confirm that TokeBi outperforms prior arts across diverse baselines and datasets.

Key facts. Cited by 5 papers in the map (data through 14 Oct 2026). Builds on 45 of its 53 references that are in the map. First posted on arXiv 29 Mar 2025.

Builds on

45 of 53 references.

References · 45

  1. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,744
  2. 2020Denoising Diffusion Probabilistic Modelscited by 21,076
  3. 2021High-Resolution Image Synthesis with Latent Diffusion Modelscited by 18,863
  4. 2019Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformercited by 14,599
  5. 2022Scalable Diffusion Models with Transformerscited by 6,054
  6. 2023SDXL: Improving Latent Diffusion Models for High-Resolution Image Synthesiscited by 4,356
  7. 2022BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generationcited by 4,322
  8. 2024Scaling Rectified Flow Transformers for High-Resolution Image Synthesiscited by 4,181
  9. 2022Prompt-to-Prompt Image Editing with Cross Attention Controlcited by 2,431
  10. 2023GenEval: An Object-Focused Framework for Evaluating Text-to-Image Alignmentcited by 972
  11. 2023PixArt-α: Fast Training of Diffusion Transformer for Photorealistic Text-to-Image Synthesiscited by 864
  12. 2023Attend-and-Excite: Attention-Based Semantic Guidance for Text-to-Image Diffusion Modelscited by 702
33 more
  1. 2024ELLA: Equip Diffusion Models with LLM for Enhanced Semantic Alignmentcited by 475
  2. 2024Evaluating Text-to-Visual Generation with Image-to-Text Generationcited by 474
  3. 2022Training-Free Structured Diffusion Guidance for Compositional Text-to-Image Synthesiscited by 388
  4. 2024PixArt-Sigma: Weak-to-Strong Training of Diffusion Transformer for 4K Text-to-Image Generationcited by 329
  5. 2023Training-Free Layout Control with Cross-Attention Guidancecited by 313
  6. 2023BoxDiff: Text-to-Image Synthesis with Training-Free Box-Constrained Diffusioncited by 282
  7. 2022Optimizing Prompts for Text-to-Image Generationcited by 207
  8. 2023VIEScore: Towards Explainable Metrics for Conditional Image Synthesis Evaluationcited by 201
  9. 2023Linguistic Binding in Diffusion Models: Enhancing Attribute Correspondence through Attention Map Alignmentcited by 171
  10. 2023Grounded Text-to-Image Synthesis with Attention Refocusingcited by 164
  11. 2023Dense Text-to-Image Generation with Attention Modulationcited by 163
  12. 2024Lumina-Next: Making Lumina-T2X Stronger and Faster with Next-DiTcited by 132
  13. 2024Training-Free Consistent Text-to-Image Generationcited by 125
  14. 2014Word Representations via Gaussian Embeddingcited by 96
  15. 2023Compositional Text-to-Image Synthesis with Attention Map Control of Diffusion Modelscited by 83
  16. 2023Divide & Bind Your Attention for Improved Generative Semantic Nursingcited by 81
  17. 2023Ranni: Taming Text-to-Image Diffusion for Accurate Instruction Followingcited by 77
  18. 2023Directed Diffusion: Direct Control of Object Placement through Attention Guidancecited by 69
  19. 2024Be Yourself: Bounded Attention for Multi-Subject Text-to-Image Generationcited by 69
  20. 2024CoMat: Aligning Text-to-Image Diffusion Model with Image-to-Text Concept Matchingcited by 61
  21. 2022DALLE-2 is Seeing Double: Flaws in Word-to-Concept Mapping in Text2Image Modelscited by 51
  22. 2022Shape-Guided Diffusion with Inside-Outside Attentioncited by 45
  23. 2024Token Merging for Training-Free Semantic Binding in Text-to-Image Synthesiscited by 35
  24. 2024Attention Calibration for Disentangled Text-to-Image Personalizationcited by 31
  25. 2023SUR-adapter: Enhancing Text-to-Image Pre-trained Diffusion Models with Large Language Modelscited by 30
  26. 2020Conditional Image Generation and Manipulation for User-Specified Contentcited by 20
  27. 2024Magnet: We Never Know How Text-to-Image Diffusion Models Work, Until We Learn How Vision-Language Models Functioncited by 19
  28. 2024A Cat Is A Cat (Not A Dog!): Unraveling Information Mix-ups in Text-to-Image Encoders through Causal Analysis and Embedding Optimizationcited by 19
  29. 2023Predicated Diffusion: Predicate Logic-Based Attention Guidance for Text-to-Image Diffusion Modelscited by 16
  30. 2023Training-Free Location-Aware Text-to-Image Synthesiscited by 13
  31. 2025Padding Tone: A Mechanistic Analysis of Padding Tokens in T2I Modelscited by 13
  32. 2024Text Embedding is Not All You Need: Attention Control for Text-to-Image Semantic Alignment with Text Self-Attention Mapscited by 12
  33. 2024Attention Overlap Is Responsible for The Entity Missing Problem in Text-to-image Diffusion Models!cited by 9

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 1; 2026: 4.

5 papers.

  1. 2025Training-free Mixed-Resolution Latent Upsampling for Spatially Accelerated Diffusion Transformerscited by 10
  2. 2026Rethinking Global Text Conditioning in Diffusion Transformerscited by 0
  3. 2026Formalizing the Binding Problemcited by 0
  4. 2026STEDiff: Strengthening Text Embedding for Text-to-Image Alignment in Diffusion Modelcited by 0
  5. 2026Intermediate Text Representation Guided Text-to-Image Generation for Enhancing One-and-Only Alignmentcited by 0