mlsanityData to 14 Oct 2026

Order Is Not Layout: Order-to-Space Bias in Image Generation

arXiv 2603.03714 ↗First posted 4 Mar 2026cs.CL cs.AI cs.CVcited by 1

Not on the map. Why?

Abstract (arXiv)

We study a systematic bias in modern image generation models: the mention order of entities in text spuriously determines spatial layout and entity--role binding. We term this phenomenon Order-to-Space Bias (OTS) and show that it arises in both text-to-image and image-to-image generation, often overriding grounded cues and causing incorrect layouts or swapped assignments. To quantify OTS, we introduce OTS-Bench, which isolates order effects with paired prompts differing only in entity order and evaluates models along two dimensions: homogenization and correctness. Experiments show that Order-to-Space Bias (OTS) is widespread in modern image generation models, and provide evidence that it is primarily data-driven and manifests during the early stages of layout formation. Motivated by this insight, we show that both targeted fine-tuning and early-stage intervention strategies can substantially reduce OTS, while preserving generation quality.

Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Builds on 35 of its 47 references that are in the map. First posted on arXiv 4 Mar 2026.

Builds on

35 of 47 references.

References · 35

  1. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,744
  2. 2020Denoising Diffusion Probabilistic Modelscited by 21,076
  3. 2014Microsoft COCO: Common Objects in Contextcited by 17,608
  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. 2023Adding Conditional Control to Text-to-Image Diffusion Modelscited by 5,503
  7. 2023SDXL: Improving Latent Diffusion Models for High-Resolution Image Synthesiscited by 4,356
  8. 2024Scaling Rectified Flow Transformers for High-Resolution Image Synthesiscited by 4,181
  9. 2022LAION-5B: An open large-scale dataset for training next generation image-text modelscited by 4,152
  10. 2022InstructPix2Pix: Learning to Follow Image Editing Instructionscited by 2,611
  11. 2023IP-Adapter: Text Compatible Image Prompt Adapter for Text-to-Image Diffusion Modelscited by 1,464
  12. 2017The "something something" video database for learning and evaluating visual common sensecited by 1,103
23 more
  1. 2023ImageReward: Learning and Evaluating Human Preferences for Text-to-Image Generationcited by 1,045
  2. 2025Qwen-Image Technical Reportcited by 1,044
  3. 2023GenEval: An Object-Focused Framework for Evaluating Text-to-Image Alignmentcited by 972
  4. 2023PixArt-α: Fast Training of Diffusion Transformer for Photorealistic Text-to-Image Synthesiscited by 864
  5. 2017AVA: A Video Dataset of Spatio-temporally Localized Atomic Visual Actionscited by 449
  6. 2023TIFA: Accurate and Interpretable Text-to-Image Faithfulness Evaluation with Question Answeringcited by 428
  7. 2022Easily Accessible Text-to-Image Generation Amplifies Demographic Stereotypes at Large Scalecited by 262
  8. 2023T2I-CompBench++: An Enhanced and Comprehensive Benchmark for Compositional Text-to-image Generationcited by 260
  9. 2023Stable Bias: Analyzing Societal Representations in Diffusion Modelscited by 226
  10. 2023Fair Diffusion: Instructing Text-to-Image Generation Models on Fairnesscited by 151
  11. 2025OmniSpatial: Towards Comprehensive Spatial Reasoning Benchmark for Vision Language Modelscited by 114
  12. 2022Animal Kingdom: A Large and Diverse Dataset for Animal Behavior Understandingcited by 103
  13. 2024Survey of Bias In Text-to-Image Generation: Definition, Evaluation, and Mitigationcited by 66
  14. 2024ConceptMix: A Compositional Image Generation Benchmark with Controllable Difficultycited by 61
  15. 2023TIBET: Identifying and Evaluating Biases in Text-to-Image Generative Modelscited by 52
  16. 2024Exploring Diffusion Time-steps for Unsupervised Representation Learningcited by 36
  17. 2024SPHERE: Unveiling Spatial Blind Spots in Vision-Language Models Through Hierarchical Evaluationcited by 31
  18. 2022Exploiting Cultural Biases via Homoglyphs in Text-to-Image Synthesiscited by 29
  19. 2024ViSAGe: A Global-Scale Analysis of Visual Stereotypes in Text-to-Image Generationcited by 27
  20. 2025Identifying and Mitigating Position Bias of Multi-image Vision-Language Modelscited by 25
  21. 2024FAIntbench: A Holistic and Precise Benchmark for Bias Evaluation in Text-to-Image Modelscited by 22
  22. 2024Multilingual Text-to-Image Generation Magnifies Gender Stereotypes and Prompt Engineering May Not Help Youcited by 15
  23. 2026The Spatial Blindspot of Vision-Language Modelscited by 8

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 1.

1 paper.

  1. 2026The Plan, Not the Decoder: Diagnosing and Repairing Compositional Failure in Reasoning-Augmented Text-to-Image Generationcited by 0