← back to paper
arxiv: 2605.12213 · 2 revisions
Goal-Oriented Reasoning for RAG-based Memory in Conversational Agentic LLM Systems