一作ACL 2026Findings· Findings of ACL
Q uery L ink: Leveraging Query-Memory Alignment for Long-Term Reasoning in LLM Agents
Xuxian Hu · Zhu Teng · Wei Zhang · Ming He · Jianping Fan
摘要
Retrieval-Augmented Generation (RAG) systems are widely used to mitigate the stateless nature of Large Language Models (LLMs) in long-term and personalized interactions by incorporating external memory. However, existing approaches often prioritize memory organization, such as knowledge graphs, w...