# Scaling Properties of Text Conditioning in Visual Generation

arXiv: https://arxiv.org/abs/2607.29679 · HTML: https://mlsanity.com/paper/2607.29679/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 27 of its 64 references that are in the map.
- First posted on arXiv 31 Jul 2026.

## Builds on (most cited first, 27 of 27)

- [LoRA: Low-Rank Adaptation of Large Language Models](https://mlsanity.com/paper/2106.09685/) (2021)
- [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://mlsanity.com/paper/2402.03300/) (2024)
- [Scaling Laws for Neural Language Models](https://mlsanity.com/paper/2001.08361/) (2020)
- [SDXL: Improving Latent Diffusion Models for High-Resolution Image Synthesis](https://mlsanity.com/paper/2307.01952/) (2023)
- [SAM 2: Segment Anything in Images and Videos](https://mlsanity.com/paper/2408.00714/) (2024)
- [Qwen-Image Technical Report](https://mlsanity.com/paper/2508.02324/) (2025)
- [Emerging Properties in Unified Multimodal Pretraining](https://mlsanity.com/paper/2505.14683/) (2025)
- [ELLA: Equip Diffusion Models with LLM for Enhanced Semantic Alignment](https://mlsanity.com/paper/2403.05135/) (2024)
- [OmniGen2: Towards Instruction-Aligned Multimodal Generation](https://mlsanity.com/paper/2506.18871/) (2025)
- [WISE: A World Knowledge-Informed Semantic Evaluation for Text-to-Image Generation](https://mlsanity.com/paper/2503.07265/) (2025)
- [Cosmos 3: Omnimodal World Models for Physical AI](https://mlsanity.com/paper/2606.02800/) (2026)
- [HunyuanImage 3.0 Technical Report](https://mlsanity.com/paper/2509.23951/) (2025)
- [Emu3.5: Native Multimodal Models are World Learners](https://mlsanity.com/paper/2510.26583/) (2025)
- [Echo-4o: Harnessing the Power of GPT-4o Synthetic Images for Improved Image Generation](https://mlsanity.com/paper/2508.09987/) (2025)
- [LongCat-Image Technical Report](https://mlsanity.com/paper/2512.07584/) (2025)
- [Improving the Diffusability of Autoencoders](https://mlsanity.com/paper/2502.14831/) (2025)
- [TIIF-Bench: How Does Your T2I Model Follow Your Instructions?](https://mlsanity.com/paper/2506.02161/) (2025)
- [GenEval 2: Addressing Benchmark Drift in Text-to-Image Evaluation](https://mlsanity.com/paper/2512.16853/) (2025)
- [Scaling Laws For Diffusion Transformers](https://mlsanity.com/paper/2410.08184/) (2024)
- [PromptEnhancer: A Simple Approach to Enhance Text-to-Image Models via Chain-of-Thought Prompt Rewriting](https://mlsanity.com/paper/2509.04545/) (2025)
- [A Picture is Worth a Thousand Words: Principled Recaptioning Improves Image Generation](https://mlsanity.com/paper/2310.16656/) (2023)
- [GEMS: Agent-Native Multimodal Generation with Memory and Skills](https://mlsanity.com/paper/2603.28088/) (2026)
- [Generating an Image From 1,000 Words: Enhancing Text-to-Image With Structured Captions](https://mlsanity.com/paper/2511.06876/) (2025)
- [DetailMaster: Can Your Text-to-Image Model Handle Long Prompts?](https://mlsanity.com/paper/2505.16915/) (2025)
- [Harnessing Caption Detailness for Data-Efficient Text-to-Image Generation](https://mlsanity.com/paper/2505.15172/) (2025)
- [POET: Preference Optimization for Enhanced Text-to-Image Generation](https://mlsanity.com/paper/2510.12041/) (2025)
- [Structured Captions Improve Prompt Adherence in Text-to-Image Models (Re-LAION-Caption 19M)](https://mlsanity.com/paper/2507.05300/) (2025)

## Built on by (most cited first, 2 of 2)

- [From Corpora to Co-Evolving Capabilities: Capability-Centric Data Design for Generalist Image Generation](https://mlsanity.com/paper/2608.18076/) (2026)
- [VidaForge: Open Research Infrastructure for Video Pretraining Data Recipes](https://mlsanity.com/paper/2609.06652/) (2026)

## Abstract (arXiv)

We study empirical scaling properties for text conditioning in visual generation. Such properties have rarely been measured because diffusion loss does not scale with the number of tokens in natural-language prompts. Surprisingly, we find that the converged diffusion loss scales with the amount of structured language in the prompt. To quantify structured language, we adapt two complementary measures: a white-box likelihood metric (GPG) and a black-box attribute metric (ED). Across controlled training runs, the converged diffusion loss decreases approximately linearly with GPG and follows a power law with ED. Guided by these scaling properties, we improve diffusability by constructing structured prompts with semantic and geometric annotations derived from images, and improve promptability by training a prompter through supervised fine-tuning, cold-start, and verifier-gated on-policy distillation. The resulting system outperforms all evaluated open-weight models on nearly every compositional, reasoning, and world-knowledge benchmark, while matching or surpassing the strongest closed-weight models on most evaluations.