Attribute Token Arithmetic: Disentangled and Continuous Semantic Control for Visual Autoregressive Models

2026-08-28

AI Summary

本文提出属性令牌算术(ATA)方法，解决自回归视觉模型中细粒度语义控制问题，通过预训练潜在空间中的简单算术操作实现连续且解耦的属性调整。

Abstract

Autoregressive text-to-image generation has recently achieved remarkable progress, offering high-fidelity synthesis via a unified generative framework. However, fine-grained semantic control remains challenging due to the attribute entanglement and the misalignment between textual and fine-grained visual representations. In this paper, we introduce Attribute Token Arithmetic (ATA), a method that enables disentangled and continuous attribute control in visual autoregressive modelling. Inspired by the vector arithmetic property observed in word embeddings, ATA identifies semantic directions corresponding to visual attributes (e.g., aging, fatness, emotion) directly within the pretrained autoregressive latent space. These directions are learned from a single reference image, without model retraining or large-scale supervision. During generation, attributes can be continuously adjusted and compositionally combined through simple arithmetic operations with other attribute tokens. Extensive experiments demonstrate that ATA achieves identity-preserving, fine-grained, and multi-attribute adjustment, outperforming existing autoregressive editing baselines in controllability, generality, and computational efficiency. Our code will be available at https://github.com/Madaoer/ATA.

Problem

• autoregressive text-to-image generation
• semantic control
• attribute entanglement
• visual representation

Innovation

• Attribute Token Arithmetic
• disentangled and continuous attribute control
• visual autoregressive models