This paper explores the role of cross-attention and self-attention maps in Stable Diffusion for text-guided image editing (TIE). The authors conduct a probing analysis to understand the semantic meanings of these attention maps and their impact on image editing. They find that cross-attention maps often contain object attribution information, which can lead to editing failures, while self-attention maps play a crucial role in preserving the geometric and shape details of the source image during transformation. Based on these findings, they propose a simplified, tuning-free method called Free-Prompt-Editing (FPE), which modifies only the self-attention maps during the denoising process. Experimental results show that FPE consistently outperforms popular approaches on multiple datasets, demonstrating its effectiveness and efficiency in TIE tasks. The paper contributes to a deeper understanding of attention mechanisms in diffusion models and provides a practical solution for stable and efficient image editing.This paper explores the role of cross-attention and self-attention maps in Stable Diffusion for text-guided image editing (TIE). The authors conduct a probing analysis to understand the semantic meanings of these attention maps and their impact on image editing. They find that cross-attention maps often contain object attribution information, which can lead to editing failures, while self-attention maps play a crucial role in preserving the geometric and shape details of the source image during transformation. Based on these findings, they propose a simplified, tuning-free method called Free-Prompt-Editing (FPE), which modifies only the self-attention maps during the denoising process. Experimental results show that FPE consistently outperforms popular approaches on multiple datasets, demonstrating its effectiveness and efficiency in TIE tasks. The paper contributes to a deeper understanding of attention mechanisms in diffusion models and provides a practical solution for stable and efficient image editing.