# Token Merging for Training-Free Semantic Binding in Text-to-Image Synthesis

arXiv: https://arxiv.org/abs/2411.07132 · HTML: https://mlsanity.com/paper/2411.07132/

## Key facts

- Cited by 35 papers in the map (data through 14 Oct 2026).
- Placed in Visual token pruning in multimodal models, one of 12 directions in Efficient LLM inference and serving.
- Builds on 73 of its 84 references that are in the map.
- First posted on arXiv 11 Nov 2024.

## Placed in

- Efficient LLM inference and serving / [Visual token pruning in multimodal models](https://mlsanity.com/inference/visual-token-pruning-in-multimodal-models/)

## Builds on (most cited first, 30 of 73)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [Denoising Diffusion Probabilistic Models](https://mlsanity.com/paper/2006.11239/) (2020)
- [High-Resolution Image Synthesis with Latent Diffusion Models](https://mlsanity.com/paper/2112.10752/) (2021)
- [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023)
- [Hierarchical Text-Conditional Image Generation with CLIP Latents](https://mlsanity.com/paper/2204.06125/) (2022)
- [Adding Conditional Control to Text-to-Image Diffusion Models](https://mlsanity.com/paper/2302.05543/) (2023)
- [SDXL: Improving Latent Diffusion Models for High-Resolution Image Synthesis](https://mlsanity.com/paper/2307.01952/) (2023)
- [Distributed Representations of Words and Phrases and their Compositionality](https://mlsanity.com/paper/1310.4546/) (2013)
- [Prompt-to-Prompt Image Editing with Cross Attention Control](https://mlsanity.com/paper/2208.01626/) (2022)
- [MMDetection: Open MMLab Detection Toolbox and Benchmark](https://mlsanity.com/paper/1906.07155/) (2019)
- [T2I-Adapter: Learning Adapters to Dig out More Controllable Ability for Text-to-Image Diffusion Models](https://mlsanity.com/paper/2302.08453/) (2023)
- [Grounded Language-Image Pre-training](https://mlsanity.com/paper/2112.03857/) (2021)
- [ImageReward: Learning and Evaluating Human Preferences for Text-to-Image Generation](https://mlsanity.com/paper/2304.05977/) (2023)
- [eDiff-I: Text-to-Image Diffusion Models with an Ensemble of Expert Denoisers](https://mlsanity.com/paper/2211.01324/) (2022)
- [Blended Diffusion for Text-driven Editing of Natural Images](https://mlsanity.com/paper/2111.14818/) (2021)
- [Plug-and-Play Diffusion Features for Text-Driven Image-to-Image Translation](https://mlsanity.com/paper/2211.12572/) (2022)
- [PixArt-α: Fast Training of Diffusion Transformer for Photorealistic Text-to-Image Synthesis](https://mlsanity.com/paper/2310.00426/) (2023)
- [GLIGEN: Open-Set Grounded Text-to-Image Generation](https://mlsanity.com/paper/2301.07093/) (2023)
- [Attend-and-Excite: Attention-Based Semantic Guidance for Text-to-Image Diffusion Models](https://mlsanity.com/paper/2301.13826/) (2023)
- [Compositional Visual Generation with Composable Diffusion Models](https://mlsanity.com/paper/2206.01714/) (2022)
- [Erasing Concepts from Diffusion Models](https://mlsanity.com/paper/2303.07345/) (2023)
- [ELLA: Equip Diffusion Models with LLM for Enhanced Semantic Alignment](https://mlsanity.com/paper/2403.05135/) (2024)
- [Universal Guidance for Diffusion Models](https://mlsanity.com/paper/2302.07121/) (2023)
- [Training-Free Structured Diffusion Guidance for Compositional Text-to-Image Synthesis](https://mlsanity.com/paper/2212.05032/) (2022)
- [PixArt-Sigma: Weak-to-Strong Training of Diffusion Transformer for 4K Text-to-Image Generation](https://mlsanity.com/paper/2403.04692/) (2024)
- [Training-Free Layout Control with Cross-Attention Guidance](https://mlsanity.com/paper/2304.03373/) (2023)
- [Diffusion Self-Guidance for Controllable Image Generation](https://mlsanity.com/paper/2306.00986/) (2023)
- [BoxDiff: Text-to-Image Synthesis with Training-Free Box-Constrained Diffusion](https://mlsanity.com/paper/2307.10816/) (2023)
- [T2I-CompBench++: An Enhanced and Comprehensive Benchmark for Compositional Text-to-image Generation](https://mlsanity.com/paper/2307.06350/) (2023)
- [LLM-grounded Diffusion: Enhancing Prompt Understanding of Text-to-Image Diffusion Models with Large Language Models](https://mlsanity.com/paper/2305.13655/) (2023)

## Built on by (most cited first, 30 of 35)

- [Representation Entanglement for Generation: Training Diffusion Transformers Is Much Easier Than You Think](https://mlsanity.com/paper/2507.01467/) (2025)
- [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)
- [Does Object Binding Naturally Emerge in Large Pretrained Vision Transformers?](https://mlsanity.com/paper/2510.24709/) (2025)
- [Mod-Adapter: Tuning-Free and Versatile Multi-concept Personalization via Modulation Adapter](https://mlsanity.com/paper/2505.18612/) (2025)
- [FlashSloth: Lightning Multimodal Large Language Models via Embedded Visual Compression](https://mlsanity.com/paper/2412.04317/) (2024)
- [Anchor Token Matching: Implicit Structure Locking for Training-free AR Image Editing](https://mlsanity.com/paper/2504.10434/) (2025)
- [Text-to-Image Synthesis: A Decade Survey](https://mlsanity.com/paper/2411.16164/) (2024)
- [Free-Lunch Color-Texture Disentanglement for Stylized Image Generation](https://mlsanity.com/paper/2503.14275/) (2025)
- [Addressing Text Embedding Leakage in Diffusion-based Image Editing](https://mlsanity.com/paper/2412.04715/) (2024)
- [Geometrical Properties of Text Token Embeddings for Strong Semantic Binding in Text-to-Image Generation](https://mlsanity.com/paper/2503.23011/) (2025)
- [One-Way Ticket:Time-Independent Unified Encoder for Distilling Text-to-Image Diffusion Models](https://mlsanity.com/paper/2505.21960/) (2025)
- [Steer Away From Mode Collisions: Improving Composition In Diffusion Models](https://mlsanity.com/paper/2509.25940/) (2025)
- [FreeCond: Free Lunch in the Input Conditions of Text-Guided Inpainting](https://mlsanity.com/paper/2412.00427/) (2024)
- [Follow the Flow: On Information Flow Across Textual Tokens in Text-to-Image Models](https://mlsanity.com/paper/2504.01137/) (2025)
- [Detail++: Training-Free Detail Enhancer for T2I Diffusion Models](https://mlsanity.com/paper/2507.17853/) (2025)
- [DDiT: Dynamic Patch Scheduling for Efficient Diffusion Transformers](https://mlsanity.com/paper/2602.16968/) (2026)
- [TPSO: Training-Free Diverse Image Generation via Semantic Prompt Embedding Optimization](https://mlsanity.com/paper/2511.19811/) (2025)
- [Why Settle for Mid: A Probabilistic Viewpoint to Spatial Relationship Alignment in Text-to-image Models](https://mlsanity.com/paper/2506.23418/) (2025)
- [A Conditional Probability Framework for Compositional Zero-shot Learning](https://mlsanity.com/paper/2507.17377/) (2025)
- [What "Not" to Detect: Negation-Aware VLMs via Structured Reasoning and Token Merging](https://mlsanity.com/paper/2510.13232/) (2025)
- [Adversarial Concept Distillation for One-Step Diffusion Personalization](https://mlsanity.com/paper/2510.20512/) (2025)
- [Correlation-Weighted Multi-Reward Optimization for Compositional Generation](https://mlsanity.com/paper/2603.18528/) (2026)
- [Comp-Attn: Present-and-Align Attention for Compositional Video Generation](https://mlsanity.com/paper/2503.14428/) (2025)
- [ISAC: Training-Free Instance-to-Semantic Attention Control for Multi-Instance Generation](https://mlsanity.com/paper/2505.20935/) (2025)
- [CubistMerge: Spatial-Preserving Token Merging For Diverse ViT Backbones](https://mlsanity.com/paper/2509.21764/) (2025)
- [Delta-K: Boosting Multi-Instance Generation via Cross-Attention Augmentation](https://mlsanity.com/paper/2603.10210/) (2026)
- [SimInsert: Seamless Video Object Insertion via Regional Sparse Attention Fusion](https://mlsanity.com/paper/2605.23245/) (2026)
- [Where to Refine, When to Stop: Rethinking Redundancy via Latent Discrepancy for Efficient Visual Autoregressive Generation](https://mlsanity.com/paper/2606.00310/) (2026)
- [Formalizing the Binding Problem](https://mlsanity.com/paper/2606.03976/) (2026)
- [STEDiff: Strengthening Text Embedding for Text-to-Image Alignment in Diffusion Model](https://mlsanity.com/paper/2606.10653/) (2026)

## Abstract (arXiv)

Although text-to-image (T2I) models exhibit remarkable generation capabilities, they frequently fail to accurately bind semantically related objects or attributes in the input prompts; a challenge termed semantic binding. Previous approaches either involve intensive fine-tuning of the entire T2I model or require users or large language models to specify generation layouts, adding complexity. In this paper, we define semantic binding as the task of associating a given object with its attribute, termed attribute binding, or linking it to other related sub-objects, referred to as object binding. We introduce a novel method called Token Merging (ToMe), which enhances semantic binding by aggregating relevant tokens into a single composite token. This ensures that the object, its attributes and sub-objects all share the same cross-attention map. Additionally, to address potential confusion among main objects with complex textual prompts, we propose end token substitution as a complementary strategy. To further refine our approach in the initial stages of T2I generation, where layouts are determined, we incorporate two auxiliary losses, an entropy loss and a semantic binding loss, to iteratively update the composite token to improve the generation integrity. We conducted extensive experiments to validate the effectiveness of ToMe, comparing it against various existing methods on the T2I-CompBench and our proposed GPT-4o object binding benchmark. Our method is particularly effective in complex scenarios that involve multiple objects and attributes, which previous methods often fail to address. The code will be publicly available at urlhttps://github.com/hutaihang/ToMe.