From Retrieved Context to Evidence-Seeking Trajectories: Agentic Multimodal RAG Survey
A trajectory-level survey of Agentic Multimodal RAG, covering paper lineage, core components, benchmark gaps, grounding, verification, and evaluation.
Paper reviews, research notes, and updates from KUDoc on Document AI, multimodal understanding, and long-context reasoning.
A trajectory-level survey of Agentic Multimodal RAG, covering paper lineage, core components, benchmark gaps, grounding, verification, and evaluation.
A short review of DocAgent, focusing on its agentic design for multimodal long-context document understanding.