{"ID":2853327,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2510.16342","arxiv_id":"2510.16342","title":"Beyond Fixed Anchors: Precisely Erasing Concepts with Sibling Exclusive Counterparts","abstract":"Existing concept erasure methods for text-to-image diffusion models commonly rely on fixed anchor strategies, which often lead to critical issues such as concept re-emergence and erosion. To address this, we conduct causal tracing to reveal the inherent sensitivity of erasure to anchor selection and define Sibling Exclusive Concepts as a superior class of anchors. Based on this insight, we propose \\textbf{SELECT} (Sibling-Exclusive Evaluation for Contextual Targeting), a dynamic anchor selection framework designed to overcome the limitations of fixed anchors. Our framework introduces a novel two-stage evaluation mechanism that automatically discovers optimal anchors for precise erasure while identifying critical boundary anchors to preserve related concepts. Extensive evaluations demonstrate that SELECT, as a universal anchor solution, not only efficiently adapts to multiple erasure frameworks but also consistently outperforms existing baselines across key performance metrics, averaging only 4 seconds for anchor mining of a single concept.","short_abstract":"Existing concept erasure methods for text-to-image diffusion models commonly rely on fixed anchor strategies, which often lead to critical issues such as concept re-emergence and erosion. To address this, we conduct causal tracing to reveal the inherent sensitivity of erasure to anchor selection and define Sibling Excl...","url_abs":"https://arxiv.org/abs/2510.16342","url_pdf":"https://arxiv.org/pdf/2510.16342v1","authors":"[\"Tong Zhang\",\"Ru Zhang\",\"Jianyi Liu\",\"Zhen Yang\",\"Gongshen Liu\"]","published":"2025-10-18T04:03:27Z","proceeding":"cs.AI","tasks":"[\"cs.AI\",\"cs.CV\"]","methods":"[\"Diffusion Model\"]","has_code":false}