الملخص
Text-to-image diffusion models can generate prohibited content, which motivates concept erasure through machine unlearning. Most erasure methods intervene at the text interface, through prompt modification or localized updates to text-conditioning weights, and they are evaluated by what the model outputs for given prompts. Such evaluation cannot see what the network still encodes. Latent-space auditing, which bypasses text conditioning and probes the denoising network directly, shows that erased concepts remain recoverable from internal representations. We find that this also holds for methods built to be robust against adversarial prompts, and that the problem grows with the number of erased concepts. We propose Auditing-Aware Unlearning for Verifiable Concept Erasure in Diffusion Models (AVCE), a framework that grounds erasure in the model's latent representations. AVCE audits the embedding neighborhood of each concept and condenses the discovered vulnerable directions into an anchor at the weakest geometric point. It edits cross-attention and self-attention projections in closed form at this anchor, then fine-tunes the two pathways with pathway-level auditing losses, using orthogonal gradient projection to consolidate multiple concepts. Experiments on SD v1.5, SDXL, and Flux 1.0 across object, explicit-content, and artistic-style unlearning show that AVCE reduces attack success rates by 5.07x and improves auditing scores by 3.84x over the strongest baseline, while preserving competitive generation quality.
الكلمات المفتاحية
الموضوع
بيانات النشر
- المجلة
- غير متاح
- وصول مفتوح
- وصول مفتوح أخضر
اقتبس هذه المقالة
APA 7
Deng, K., Li, Y., Li, G., Xiao, Y., Yuan, G., Yuan, X., Hui, B., & Ma, X. (2026). No Concept Escapes the Audit: Auditing-Aware Unlearning for Verifiable Concept Erasure in Diffusion Models. https://omanscience.com/ar/articles/no-concept-escapes-the-audit-auditing-aware-unlearning-for-verifiable-concept-erasure-in-diffusion-models
MLA 9
Deng, Kaiyuan, et al. "No Concept Escapes the Audit: Auditing-Aware Unlearning for Verifiable Concept Erasure in Diffusion Models." https://omanscience.com/ar/articles/no-concept-escapes-the-audit-auditing-aware-unlearning-for-verifiable-concept-erasure-in-diffusion-models.
شيكاغو (المؤلف–التاريخ)
Deng, Kaiyuan, Yuchen Li, Gen Li, Yang Xiao, Geng Yuan, Xiaoyong Yuan, Bo Hui, and Xiaolong Ma. 2026. "No Concept Escapes the Audit: Auditing-Aware Unlearning for Verifiable Concept Erasure in Diffusion Models." https://omanscience.com/ar/articles/no-concept-escapes-the-audit-auditing-aware-unlearning-for-verifiable-concept-erasure-in-diffusion-models.
هارفارد
Deng, K., Li, Y., Li, G., Xiao, Y., Yuan, G., Yuan, X., Hui, B. and Ma, X. (2026) 'No Concept Escapes the Audit: Auditing-Aware Unlearning for Verifiable Concept Erasure in Diffusion Models', Available at: https://omanscience.com/ar/articles/no-concept-escapes-the-audit-auditing-aware-unlearning-for-verifiable-concept-erasure-in-diffusion-models.
فانكوفر
Deng K, Li Y, Li G, Xiao Y, Yuan G, Yuan X, et al. No Concept Escapes the Audit: Auditing-Aware Unlearning for Verifiable Concept Erasure in Diffusion Models. https://omanscience.com/ar/articles/no-concept-escapes-the-audit-auditing-aware-unlearning-for-verifiable-concept-erasure-in-diffusion-models
IEEE
K. Deng, Y. Li, G. Li, Y. Xiao, G. Yuan, X. Yuan, B. Hui, and X. Ma, "No Concept Escapes the Audit: Auditing-Aware Unlearning for Verifiable Concept Erasure in Diffusion Models," https://omanscience.com/ar/articles/no-concept-escapes-the-audit-auditing-aware-unlearning-for-verifiable-concept-erasure-in-diffusion-models.