ReDeEP: Detecting Hallucination in Retrieval-Augmented Generation via Mechanistic Interpretability

2025-01-22 · ICLR 2025 Spotlight · anchor

Findings