CrAM: Credibility-Aware Attention Modification in LLMs for Combating Misinformation in RAG.

Boyi Deng, Wenjie Wang, Fengbin Zhu, Qifan Wang, Fuli Feng

Browse the full AAAI paper archive.