AgenticRag-R1：基于栈内存的智能体强化学习实现多步推理与检索
原标题：AgenticRag-R1: Agentic Reinforcement Learning with Stack Memory for Multi-Step Reasoning, Retrieval and Memorizing
AI 摘要
AgenticRag-R1 是一个基于强化学习的框架，通过内存栈和细粒度动作空间深度融合推理、检索和记忆，并采用分层动作感知奖励和信息感知轨迹拒绝策略，以解决现有 RAG 系统在多步推理中自适应检索和上下文修订的不足。实验表明，该方法在多个多跳、开放域和智能体推理基准上优于强基线，并展现出更稳健、可解释且记忆感知的推理行为。代码已在 GitHub 上匿名公开。
正文节选
AgenticRag-R1: Agentic Reinforcement Learning with Stack Memory for Multi-Step Reasoning, Retrieval and Memorizing Abstract Retrieval-Augmented Generation (RAG) improves the factuality of large language models (LLMs), yet existing RAG systems often struggle with complex, multi-step reasoning that requires adaptive retrieval and continuous revision of intermediate contexts. Recent reinforcement learning (RL)-based agentic RAG methods partially alleviate this issue, but typically rely on coarse-gr