Knowledge-Empowered,Collaborative,and Co-Evolving AI Models:The Post-LLM Roadmap

作     者：Fei Wu Tao Shen Thomas Bäck Jingyuan Chen Gang Huang Yaochu Jin Kun Kuang Mengze Li Cewu Lu Jiaxu Miao Yongwei Wang Ying Wei Fan Wu Junchi Yan Hongxia Yang Yi Yang Shengyu Zhang Zhou Zhao Yueting Zhuang Yunhe Pan Fei Wu;Tao Shen;Thomas Bäck;Jingyuan Chen;Gang Huang;Yaochu Jin;Kun Kuang;Mengze Li;Cewu Lu;Jiaxu Miao;Yongwei Wang;Ying Wei;Fan Wu;Junchi Yan;Hongxia Yang;Yi Yang;Shengyu Zhang;Zhou Zhao;Yueting Zhuang;Yunhe Pan

作者机构：College of Computer Science and TechnologyZhejiang UniversityHangzhou 310027China Leiden Institute of Advanced Computer ScienceLeiden UniversityLeiden 2333 CCNetherlands College of Electrical EngineeringZhejiang UniversityHangzhou 310027China School of EngineeringWestlake UniversityHangzhou 310024China Department of Computer Science and EngineeringHong Kong University of Science and TechnologyHong Kong 999077China Department of Computer Science and EngineeringShanghai Jiao Tong UniversityShanghai 200240China School of Cyber Science and TechnologySun Yat-Sen UniversityShenzhen 518107China Department of ComputingThe Hong Kong Polytechnic UniversityHong Kong 999077China 

出 版 物：《工程（英文）》 (Engineering)

年 卷 期：2025年第44卷第1期

页      面：87-100页

核心收录：

基　　金：国家自然科学基金 

主　　题：Artificial intelligence Large language models Knowledge empowerment Model collaboration Model co-evolution 

摘      要：Large language models(LLMs)have significantly advanced artificial intelligence(AI)by excelling in tasks such as understanding,generation,and reasoning across multiple *** these achieve-ments,LLMs have inherent limitations including outdated information,hallucinations,inefficiency,lack of interpretability,and challenges in domain-specific *** address these issues,this survey explores three promising directions in the post-LLM era:knowledge empowerment,model collaboration,and model ***,we examine methods of integrating external knowledge into LLMs to enhance factual accuracy,reasoning capabilities,and interpretability,including incorporating knowledge into training objectives,instruction tuning,retrieval-augmented inference,and knowledge ***,we discuss model collaboration strategies that leverage the complementary strengths of LLMs and smaller models to improve efficiency and domain-specific performance through techniques such as model merging,functional model collaboration,and knowledge ***,we delve into model co-evolution,in which multiple models collaboratively evolve by sharing knowledge,parameters,and learning strategies to adapt to dynamic environments and tasks,thereby enhancing their adaptability and continual *** illustrate how the integration of these techniques advances AI capabilities in science,engineering,and society—particularly in hypothesis development,problem formulation,problem-solving,and interpretability across various *** conclude by outlining future pathways for further advancement and applications.