Scaling Large Language Models (LLMs) typically relies on increasing the number of parameters or test-time computations to boost performance. However, these strategies are impractical for edge device deployment due to limited RAM and NPU resources. Despite hardware constraints, deploying performant LLM on edge devices such as smartphone remains crucial for user experience. To address this, we propose MeKi (Memory-based Expert Knowledge Injection), a novel system that scales LLM capacity via storage space rather than FLOPs. MeKi equips each Transformer layer with token-level memory experts that injects pre-stored semantic knowledge into the generation process. To bridge the gap between training capacity and inference efficiency, we employ a re-parameterization strategy to fold parameter matrices used during training into a compact static lookup table. By offloading the knowledge to ROM, MeKi decouples model capacity from computational cost, introducing zero inference latency overhead. Extensive experiments demonstrate that MeKi significantly outperforms dense LLM baselines with identical inference speed, validating the effectiveness of memory-based scaling paradigm for on-device LLMs. Project homepage is at https://github.com/ningding-o/MeKi.
Deep Expert Injection for Anchoring Retinal VLMs with Domain-Specific Knowledge
Beijing Institute of Technology, Beijing, China(北京理工大学);National University of Singapore, Singapore(新加坡国立大学);Tsinghua University, Beijing, China(清华大学);The Hong Kong Polytechnic University, Hong Kong(香港理工大学)
Investigating Cross-Modal Skill Injection: Scenarios, Methods, and Hyperparameters
State Key Laboratory for Multimedia Information Processing, School of Computer Science, Peking University(多媒体信息处理国家重点实验室,计算机科学学院,北京大学);WeChat AI, Tencent Inc., China(腾讯公司,中国);The University of Hong Kong(香港大学)
Generation-Augmented Generation: A Plug-and-Play Framework for Private Knowledge Injection in Large Language Models
MAIS, Institute of Automation, Chinese Academy of Sciences(自动化研究所,中国科学院);School of Artificial Intelligence, University of Chinese Academy of Sciences(中国科学院大学人工智能学院);Zhongguancun Academy(中关村学院);School of Advanced Interdisciplinary Sciences, University of Chinese Academy of Sciences(中国科学院大学先进交叉学科学院);AI Research(360人工智能研究院)
SOTOPIA-$Ω$: Dynamic Strategy Injection Learning and Social Instruction Following Evaluation for Social Agents
Institute of Information Engineering, Chinese Academy of Sciences(中国科学院信息工程研究所);School of Cyber Security, University of Chinese Academy of Sciences(中国科学院大学网络空间安全学院)
RADAR: Enhancing Radiology Report Generation with Supplementary Knowledge Injection
Department of Computing, The Hong Kong Polytechnic University(香港理工大学计算机系);Research Institute of Trustworthy Autonomous Systems and Department of Computer Science and Engineering, Southern University of Science and Technology(南方科技大学可信自主系统研究院和计算机科学与工程系);School of Computer Science, University of Nottingham Ningbo China(宁波大学计算机学院)
KAER: A Knowledge Augmented Pre-Trained Language Model for Entity Resolution
School of Information Sciences, University of Illinois Urbana-Champaign(伊利诺伊大学厄巴纳-香槟分校信息科学学院);Informatics, University of Illinois Urbana-Champaign(伊利诺伊大学厄巴纳-香槟分校信息学系)
PALoRA: Projection-Adaptive LoRA for Preserving Reasoning in Large Language Models
IT Group, Research, BNP Paribas(BNP巴黎研究院IT部门);LTCI, Télécom Paris, Institut Polytechnique de Paris(巴黎理工大学LTCI研究所);AI Institute, University of Waikato(瓦卡托大学人工智能研究所);Independent Researcher(独立研究员)
LLM-based Question-Answer Framework for Sensor-driven HVAC System Interaction
School of Integrated Technology, Yonsei University(延世大学集成技术学院);LG Electronics(LG电子);Department of Electrical and Computer Engineering, University of Michigan(密歇根大学电气与计算机工程系);Graduate School of Artificial Intelligence, POSTECH(POSTECH人工智能研究生院)
Light Alignment Improves LLM Safety via Model Self-Reflection with a Single Neuron
Beijing Institute of AI Safety and Governance(北京人工智能安全与治理研究院);Beijing Key Laboratory of Safe AI and Superalignment(北京安全人工智能与超对齐重点实验室);Ant Group Co., Ltd.(蚂蚁集团有限公司);BrainCog Lab., CASIA(CASIA脑认知实验室);Zhongguancun Academy(中关村学院)
Injecting Domain-Specific Knowledge into Large Language Models: A Comprehensive Survey
Mohamed bin Zayed University of Artificial Intelligence(Mohamed bin Zayed大学);Northeastern University(东北大学);Gaoling School of Artificial Intelligence, Renmin University of China(中国人民大学人工智能学院);ByteDance(字节跳动);Wuhan University(武汉大学)
DRAK: Unlocking Molecular Insights with Domain-Specific Retrieval-Augmented Knowledge in LLMs
Xiongan Institute of Innovation, Chinese Academy of Sciences(中国科学院雄安创新研究院);College of Computer Science and Technology, Zhejiang University(浙江大学计算机科学与技术学院);Institute of Computing Technology, Chinese Academy of Sciences(中国科学院计算技术研究所);University of Chinese Academy of Science(中国科学院大学)
Continuous Behavioral Authentication via Multi-Expert BERT Log Analysis for Secure Data Sharing
Dept. of Electrical and Computer Engineering, University of Thessaly, Greece(希腊塞萨洛尼基大学电气与计算机工程系);Centre for Research and Technology Hellas, CERTH, Greece(希腊研究中心与技术 Hellas 研究中心)