# Self-Attention Decomposition For Training Free Diffusion Editing Source: [Self-Attention Decomposition For Training Free Diffusion Editing]() Collection index: [llms.txt]() Data updated: 2026-02-14T05:00:14.824Z Authors: Tharun Anand, Mohammad Hassan Vali, Arno Solin, Green Rosh, BH Pawan Prasad Published: 2025-10-26T12:22:56.000Z ## Abstract Diffusion models achieve remarkable fidelity in image synthesis, yet precise control over their outputs for targeted editing remains challenging. A key step toward controllability is to identify interpretable directions in the model's latent representations that correspond to semantic attributes. Existing approaches for finding interpretable directions typically rely on sampling large sets of images or training auxiliary networks, which limits efficiency. We propose an analytical method that derives semantic editing directions directly from the pretrained parameters of diffusion models, requiring neither additional data nor fine-tuning. Our insight is that self-attention weight matrices encode rich structural information about the data distribution learned during training. By computing the eigenvectors of these weight matrices, we obtain robust and interpretable editing directions. Experiments demonstrate that our method produces high-quality edits across multiple datasets while reducing editing time significantly by 60% over current benchmarks. ## Recorded benchmark results A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. - [Image Editing on CelebA-HQ (test)](): Image Editing; CelebA-HQ (test); SSIM: 0.92. - [Semantic Image Editing on CelebA-HQ (test)](): Semantic Image Editing; CelebA-HQ (test); Editing Time (s): 3. ## Original sources and implementation [Original paper]() [PDF]()