2025/10/06 by Ronen Kamenetsky, Kamenetsky, Ronen, Dorfman, Sara +8 · 1 citation
Computer Science · #Artificial Intelligence (cs.AI) #Autoencoder #Computer Graphics and Visualization Techniques #Computer Vision and Pattern Recognition (cs.CV) #Control (management) #Encoder #FOS: Computer and information sciences #Generative Adversarial Networks and Image Synthesis #Graphics (cs.GR) #Image (mathematics) #Image editing #Space (punctuation)
paper · pdf · doi:10.48550/arxiv.2510.05081
published in arXiv (Cornell University) (Cornell University)
openalex publication_date 2025/10/06 · openalex created_date 2025/10/09 · openalex updated_date 2026/07/28
Large-scale text-to-image diffusion models have become the backbone of modern image editing, yet text prompts alone do not offer adequate control over the editing process. Two properties are especially desirable: disentanglement, where changing one attribute does not unintentionally alter others, and continuous control, where the strength of an edit can be smoothly adjusted. We introduce a method for disentangled and continuous editing through token-level manipulation of text embeddings. The edits are applied by manipulating the embeddings along carefully chosen directions, which control the strength of the target attribute. To identify such directions, we employ a Sparse Autoencoder (SAE), whose sparse latent space exposes semantically isolated dimensions. Our method operates directly on text embeddings without modifying the diffusion process, making it model agnostic and broadly applicable to various image synthesis backbones. Experiments show that it enables intuitive and efficient manipulations with continuous control across diverse attributes and domains.