# A Cat Is A Cat (Not A Dog!): Unraveling Information Mix-ups in Text-to-Image Encoders through Causal Analysis and Embedding Optimization

arXiv: https://arxiv.org/abs/2410.00321 · HTML: https://mlsanity.com/paper/2410.00321/

## Key facts

- Cited by 19 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 19 of its 22 references that are in the map.
- First posted on arXiv 1 Oct 2024.

## Placed in

- Diffusion models / [Image generation, text-to-image and editing](https://mlsanity.com/diffusion/image-generation-text-to-image-and-editing/)

## Builds on (most cited first, 19 of 19)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [High-Resolution Image Synthesis with Latent Diffusion Models](https://mlsanity.com/paper/2112.10752/) (2021)
- [Microsoft COCO: Common Objects in Context](https://mlsanity.com/paper/1405.0312/) (2014)
- [Hierarchical Text-Conditional Image Generation with CLIP Latents](https://mlsanity.com/paper/2204.06125/) (2022)
- [Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding](https://mlsanity.com/paper/2205.11487/) (2022)
- [SDXL: Improving Latent Diffusion Models for High-Resolution Image Synthesis](https://mlsanity.com/paper/2307.01952/) (2023)
- [BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation](https://mlsanity.com/paper/2201.12086/) (2022)
- [Scaling Rectified Flow Transformers for High-Resolution Image Synthesis](https://mlsanity.com/paper/2403.03206/) (2024)
- [Prompt-to-Prompt Image Editing with Cross Attention Control](https://mlsanity.com/paper/2208.01626/) (2022)
- [Adversarial Diffusion Distillation](https://mlsanity.com/paper/2311.17042/) (2023)
- [Attend-and-Excite: Attention-Based Semantic Guidance for Text-to-Image Diffusion Models](https://mlsanity.com/paper/2301.13826/) (2023)
- [ELLA: Equip Diffusion Models with LLM for Enhanced Semantic Alignment](https://mlsanity.com/paper/2403.05135/) (2024)
- [Training-Free Structured Diffusion Guidance for Compositional Text-to-Image Synthesis](https://mlsanity.com/paper/2212.05032/) (2022)
- [Scaling Open-Vocabulary Object Detection](https://mlsanity.com/paper/2306.09683/) (2023)
- [Linguistic Binding in Diffusion Models: Enhancing Attribute Correspondence through Attention Map Alignment](https://mlsanity.com/paper/2306.08877/) (2023)
- [Grounded Text-to-Image Synthesis with Attention Refocusing](https://mlsanity.com/paper/2306.05427/) (2023)
- [Towards Understanding Cross and Self-Attention in Stable Diffusion for Text-Guided Image Editing](https://mlsanity.com/paper/2403.03431/) (2024)
- [Predicated Diffusion: Predicate Logic-Based Attention Guidance for Text-to-Image Diffusion Models](https://mlsanity.com/paper/2311.16117/) (2023)

## Built on by (most cited first, 19 of 19)

- [Token Merging for Training-Free Semantic Binding in Text-to-Image Synthesis](https://mlsanity.com/paper/2411.07132/) (2024)
- [T2I-Copilot: A Training-Free Multi-Agent Text-to-Image System for Enhanced Prompt Interpretation and Interactive Generation](https://mlsanity.com/paper/2507.20536/) (2025)
- [The Devil is in the Prompts: Retrieval-Augmented Prompt Optimization for Text-to-Video Generation](https://mlsanity.com/paper/2504.11739/) (2025)
- [RAPO++: Cross-Stage Prompt Optimization for Text-to-Video Generation via Data Alignment and Test-Time Scaling](https://mlsanity.com/paper/2510.20206/) (2025)
- [Text Slider: Efficient and Plug-and-Play Continuous Concept Control for Image/Video Synthesis via LoRA Adapters](https://mlsanity.com/paper/2509.18831/) (2025)
- [Geometrical Properties of Text Token Embeddings for Strong Semantic Binding in Text-to-Image Generation](https://mlsanity.com/paper/2503.23011/) (2025)
- [FreeCond: Free Lunch in the Input Conditions of Text-Guided Inpainting](https://mlsanity.com/paper/2412.00427/) (2024)
- [Detail++: Training-Free Detail Enhancer for T2I Diffusion Models](https://mlsanity.com/paper/2507.17853/) (2025)
- [MapReduce LoRA: Advancing the Pareto Front in Multi-Preference Optimization for Generative Models](https://mlsanity.com/paper/2511.20629/) (2025)
- [Localized Concept Erasure in Text-to-Image Diffusion Models via High-Level Representation Misdirection](https://mlsanity.com/paper/2602.19631/) (2026)
- [Not All Parameters Matter: Masking Diffusion Models for Enhancing Generation Ability](https://mlsanity.com/paper/2505.03097/) (2025)
- [DOS: Directional Object Separation in Text Embeddings for Multi-Object Image Generation](https://mlsanity.com/paper/2510.14376/) (2025)
- [Attention, May I Have Your Decision? Localizing Generative Choices in Diffusion Models](https://mlsanity.com/paper/2604.06052/) (2026)
- [Diagnosing and Correcting Concept Omission in Multimodal Diffusion Transformers](https://mlsanity.com/paper/2605.14270/) (2026)
- [Neural Antidote: Class-Wise Prompt Tuning for Purifying Backdoors in CLIP](https://mlsanity.com/paper/2502.19269/) (2025)
- [SPF-Portrait: Towards Pure Text-to-Portrait Customization with Semantic Pollution-Free Fine-Tuning](https://mlsanity.com/paper/2504.00396/) (2025)
- [ISAC: Training-Free Instance-to-Semantic Attention Control for Multi-Instance Generation](https://mlsanity.com/paper/2505.20935/) (2025)
- [STEDiff: Strengthening Text Embedding for Text-to-Image Alignment in Diffusion Model](https://mlsanity.com/paper/2606.10653/) (2026)
- [DR.WILSS: Diffusion-Based Replay for Weakly Supervised Continual Semantic Segmentation](https://mlsanity.com/paper/2609.18444/) (2026)

## Abstract (arXiv)

This paper analyzes the impact of causal manner in the text encoder of text-to-image (T2I) diffusion models, which can lead to information bias and loss. Previous works have focused on addressing the issues through the denoising process. However, there is no research discussing how text embedding contributes to T2I models, especially when generating more than one object. In this paper, we share a comprehensive analysis of text embedding: i) how text embedding contributes to the generated images and ii) why information gets lost and biases towards the first-mentioned object. Accordingly, we propose a simple but effective text embedding balance optimization method, which is training-free, with an improvement of 125.42% on information balance in stable diffusion. Furthermore, we propose a new automatic evaluation metric that quantifies information loss more accurately than existing methods, achieving 81% concordance with human assessments. This metric effectively measures the presence and accuracy of objects, addressing the limitations of current distribution scores like CLIP's text-image similarities.