# TokenAR: Multiple Subject Generation via Autoregressive Token-level enhancement

arXiv: https://arxiv.org/abs/2510.16332 · HTML: https://mlsanity.com/paper/2510.16332/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 31 of its 34 references that are in the map.
- First posted on arXiv 18 Oct 2025.

## Builds on (most cited first, 30 of 31)

- [Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer](https://mlsanity.com/paper/1910.10683/) (2019)
- [Neural Discrete Representation Learning](https://mlsanity.com/paper/1711.00937/) (2017)
- [DreamBooth: Fine Tuning Text-to-Image Diffusion Models for Subject-Driven Generation](https://mlsanity.com/paper/2208.12242/) (2022)
- [SAM 2: Segment Anything in Images and Videos](https://mlsanity.com/paper/2408.00714/) (2024)
- [Generating Diverse High-Fidelity Images with VQ-VAE-2](https://mlsanity.com/paper/1906.00446/) (2019)
- [Scaling Autoregressive Models for Content-Rich Text-to-Image Generation](https://mlsanity.com/paper/2206.10789/) (2022)
- [Visual Autoregressive Modeling: Scalable Image Generation via Next-Scale Prediction](https://mlsanity.com/paper/2404.02905/) (2024)
- [Autoregressive Model Beats Diffusion: Llama for Scalable Image Generation](https://mlsanity.com/paper/2406.06525/) (2024)
- [Vector-quantized Image Modeling with Improved VQGAN](https://mlsanity.com/paper/2110.04627/) (2021)
- [AnyDoor: Zero-shot Object-level Image Customization](https://mlsanity.com/paper/2307.09481/) (2023)
- [OmniGen: Unified Image Generation](https://mlsanity.com/paper/2409.11340/) (2024)
- [Hard Prompts Made Easy: Gradient-Based Discrete Optimization for Prompt Tuning and Discovery](https://mlsanity.com/paper/2302.03668/) (2023)
- [Infinity: Scaling Bitwise AutoRegressive Modeling for High-Resolution Image Synthesis](https://mlsanity.com/paper/2412.04431/) (2024)
- [UniControl: A Unified Diffusion Model for Controllable Visual Generation In the Wild](https://mlsanity.com/paper/2305.11147/) (2023)
- [MS-Diffusion: Multi-subject Zero-shot Image Personalization with Layout Guidance](https://mlsanity.com/paper/2406.07209/) (2024)
- [ACE++: Instruction-Based Image Creation and Editing via Context-Aware Content Filling](https://mlsanity.com/paper/2501.02487/) (2025)
- [Insert Anything: Image Insertion via In-Context Editing in DiT](https://mlsanity.com/paper/2504.15009/) (2025)
- [Deep Human Parsing with Active Template Regression](https://mlsanity.com/paper/1503.02391/) (2015)
- [Resolving Multi-Condition Confusion for Finetuning-Free Personalized Image Generation](https://mlsanity.com/paper/2409.17920/) (2024)
- [FitDiT: Advancing the Authentic Garment Details for High-fidelity Virtual Try-on](https://mlsanity.com/paper/2411.10499/) (2024)
- [EditAR: Unified Conditional Generation with Autoregressive Models](https://mlsanity.com/paper/2501.04699/) (2025)
- [Ctrl-X: Controlling Structure and Appearance for Text-To-Image Generation Without Guidance](https://mlsanity.com/paper/2406.07540/) (2024)
- [UniCombine: Unified Multi-Conditional Combination with Diffusion Transformer](https://mlsanity.com/paper/2503.09277/) (2025)
- [Object-level Visual Prompts for Compositional Image Generation](https://mlsanity.com/paper/2501.01424/) (2025)
- [DiffuMatting: Synthesizing Arbitrary Objects with Matting-level Annotation](https://mlsanity.com/paper/2403.06168/) (2024)
- [VTON-HandFit: Virtual Try-on for Arbitrary Hand Pose Guided by Hand Priors Embedding](https://mlsanity.com/paper/2408.12340/) (2024)
- [DreamRelation: Bridging Customization and Relation Generation](https://mlsanity.com/paper/2410.23280/) (2024)
- [Continuous Visual Autoregressive Generation via Score Maximization](https://mlsanity.com/paper/2505.07812/) (2025)
- [BEN: Using Confidence-Guided Matting for Dichotomous Image Segmentation](https://mlsanity.com/paper/2501.06230/) (2025)
- [Context-Aware Autoregressive Models for Multi-Conditional Image Generation](https://mlsanity.com/paper/2505.12274/) (2025)

## Abstract (arXiv)

Autoregressive Model (AR) has shown remarkable success in conditional image generation. However, these approaches for multiple reference generation struggle with decoupling different reference identities. In this work, we propose the TokenAR framework, specifically focused on a simple but effective token-level enhancement mechanism to address reference identity confusion problem. Such token-level enhancement consists of three parts, 1). Token Index Embedding clusters the tokens index for better representing the same reference images; 2). Instruct Token Injection plays as a role of extra visual feature container to inject detailed and complementary priors for reference tokens; 3). The identity-token disentanglement strategy (ITD) explicitly guides the token representations toward independently representing the features of each identity.This token-enhancement framework significantly augments the capabilities of existing AR based methods in conditional image generation, enabling good identity consistency while preserving high quality background reconstruction. Driven by the goal of high-quality and high-diversity in multi-subject generation, we introduce the InstructAR Dataset, the first open-source, large-scale, multi-reference input, open domain image generation dataset that includes 28K training pairs, each example has two reference subjects, a relative prompt and a background with mask annotation, curated for multiple reference image generation training and evaluating. Comprehensive experiments validate that our approach surpasses current state-of-the-art models in multiple reference image generation task. The implementation code and datasets will be made publicly. Codes are available, see https://github.com/lyrig/TokenAR