AI RESEARCH
Memorization In Stable Diffusion Is Unexpectedly Driven by CLIP Embeddings
arXiv CS.LG
•
ArXi:2605.02908v1 Announce Type: cross Understanding how textual embeddings contribute to memorization in text-to-image diffusion models is crucial for both interpretability and safety. This paper investigates an unexpected behavior of CLIP embeddings in Stable Diffusion, revealing that the model disproportionately relies on specific embeddings. We categorize input tokens as,, and with corresponding embeddings $\mathbf{}^{\mathbf{sot}}, \mathbf{}^{\mathbf{pr}}, \mathbf{}^{\mathbf{eot}}, \mathbf{}^{\mathbf{pad.