LATIDIA · Ciberseguridad

Exponer la ilusión del borrado en la edición de conocimientos para LLM

arXiv:2606.23276v3 Announce Type: replace-cross Abstract: Knowledge Editing (KE) se ha convertido en una frontera para actualizar hechos específicos en LLM sin un costoso reentrenamiento, pero su confiabilidad y mecanismos subyacentes permanecen

La noticia

arXiv:2606.23276v3 Announce Type: replace-cross Abstract: Knowledge Editing (KE) has emerged as a frontier for updating specific facts in LLMs without costly retraining, but its reliability and underlying mechanisms remain poorly understood. In this work, we examine KE from an adversarial elicitation perspective, revealing that edited knowledge is often not fully erased and continues to surface, with consistent failures observed across diverse model architectures. To explain this behavior, we conduct a mechanistic analysis of popular KE methods. We show that low-rank updates do not overwrite existing knowledge but instead redistribute it within the model's representation space. Furthermore, we find that these methods act as targeted suppression mechanisms that reduce the likelihood