MODE-RAG: Manifold Outlier Diagnosis and Energy-based Retrieval-Augmented Generation Evaluation
MODE-RAG proposes a multi-agent system driven by Variational Free Energy and Monte Carlo Tree Search to dynamically gate and guide interventions in multimodal RAG. It establishes a principled energy-based reasoning loop to quantify and mitigate complex failure modes like cross-modal hallucinations and sycophancy in generative models.
Short on time? Watch the video overview of this paper.
Listen on: Spotify · Apple Podcasts · Amazon Music · YouTube
Follow the show: YouTube · Medium · X · LinkedIn · Bluesky · Threads · Instagram · TikTok · Facebook · Telegram