LLM Alignment-Utility Asymmetry under Semantic-Preserving TransformationsSep 1, 2026·Mohan Li,Chengyu YuFrancesco SovranoMarc LangheinrichMartin Gjoreski· 0 min read URL Cite DOITypeJournal articlePublicationCoRRpublicationsLast updated on Sep 1, 2026 AuthorsMohan LiPhD StudentFrancesco SovranoPostdocMarc LangheinrichProfessorMartin GjoreskiPostdoc← Explanation Navigator: Rectifying Out-of-Scope Human Interpretations of Leaky AI Explanations through Conversational Guidance Sep 1, 2026Can Global XAI Methods Reveal Injected Behaviours in LLMs? SHAP vs Rule Extraction vs RuleSHAP Aug 8, 2026 →
