%% Bibliography for the numeric-citation ACM version. %% Additional sources cited in Sections 1 and 2 (2026-09-16). @inproceedings{yao2023react, author = {Yao, Shunyu and Zhao, Jeffrey and Yu, Dian and Du, Nan and Shafran, Izhak and Narasimhan, Karthik and Cao, Yuan}, title = {{ReAct: Synergizing Reasoning and Acting in Language Models}}, booktitle = {The Eleventh International Conference on Learning Representations}, year = {2023}, url = {https://openreview.net/forum?id=WE_vluYUL-X} } @misc{hu2025memoryage, author = {Hu, Yuyang and Liu, Shichun and Yue, Yanwei and Zhang, Guibin and Liu, Boyang and Zhu, Fangyi and Lin, Jiahang and Guo, Honglin and Dou, Shihan and Xi, Zhiheng and Jin, Senjie and Tan, Jiejun and Yin, Yanbin and Liu, Jiongnan and Zhang, Zeyu and Sun, Zhongxiang and Zhu, Yutao and Sun, Hao and Peng, Boci and Cheng, Zhenrong and Fan, Xuanbo and Guo, Jiaxin and Yu, Xinlei and Zhou, Zhenhong and Hu, Zewen and Huo, Jiahao and Wang, Junhao and Niu, Yuwei and Wang, Yu and Yin, Zhenfei and Hu, Xiaobin and Liao, Yue and Li, Qiankun and Wang, Kun and Zhou, Wangchunshu and Liu, Yixin and Cheng, Dawei and Zhang, Qi and Gui, Tao and Pan, Shirui and Zhang, Yan and Torr, Philip and Dou, Zhicheng and Wen, Ji-Rong and Huang, Xuanjing and Jiang, Yu-Gang and Yan, Shuicheng}, title = {{Memory in the Age of AI Agents}}, howpublished = {arXiv preprint arXiv:2512.13564}, year = {2025}, url = {https://arxiv.org/abs/2512.13564} } @inproceedings{shinn2023reflexion, author = {Shinn, Noah and Cassano, Federico and Gopinath, Ashwin and Narasimhan, Karthik and Yao, Shunyu}, title = {{Reflexion: Language Agents with Verbal Reinforcement Learning}}, booktitle = {Advances in Neural Information Processing Systems}, volume = {36}, pages = {8634--8652}, year = {2023}, url = {https://proceedings.neurips.cc/paper_files/paper/2023/hash/1b44b878bb782e6954cd888628510e90-Abstract-Conference.html} } @article{zhou2019edgeintelligence, author = {Zhou, Zhi and Chen, Xu and Li, En and Zeng, Liekang and Luo, Ke and Zhang, Junshan}, title = {{Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing}}, journal = {Proceedings of the IEEE}, volume = {107}, number = {8}, pages = {1738--1762}, year = {2019}, doi = {10.1109/JPROC.2019.2918951} } @article{wang2025empoweringedge, author = {Wang, Xubin and Tang, Zhiqing and Guo, Jianxiong and Meng, Tianhui and Wang, Chenhao and Wang, Tian and Jia, Weijia}, title = {{Empowering Edge Intelligence: A Comprehensive Survey on On-Device AI Models}}, journal = {ACM Computing Surveys}, volume = {57}, number = {9}, pages = {1--39}, year = {2025}, doi = {10.1145/3724420} } @inproceedings{zulfikar2024memoro, author = {Zulfikar, Wazeer Deen and Chan, Samantha W. T. and Maes, Pattie}, title = {{Memoro: Using Large Language Models to Realize a Concise Interface for Real-Time Memory Augmentation}}, booktitle = {Proceedings of the CHI Conference on Human Factors in Computing Systems}, articleno = {450}, numpages = {18}, publisher = {ACM}, year = {2024}, doi = {10.1145/3613904.3642450} } @inproceedings{alizadeh2024llmflash, author = {Alizadeh, Keivan and Mirzadeh, Seyed Iman and Belenko, Dmitry and Khatamifard, S. and Cho, Minsik and Del Mundo, Carlo C. and Rastegari, Mohammad and Farajtabar, Mehrdad}, title = {{LLM in a flash: Efficient Large Language Model Inference with Limited Memory}}, booktitle = {Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)}, pages = {12562--12584}, publisher = {Association for Computational Linguistics}, year = {2024}, doi = {10.18653/v1/2024.acl-long.678} } @article{zhong2024memorybank, author = {Zhong, Wanjun and Guo, Lianghong and Gao, Qiqi and Ye, He and Wang, Yanlin}, title = {{MemoryBank: Enhancing Large Language Models with Long-Term Memory}}, journal = {Proceedings of the AAAI Conference on Artificial Intelligence}, volume = {38}, number = {17}, pages = {19724--19731}, year = {2024}, doi = {10.1609/aaai.v38i17.29946} } @inproceedings{salemi2024lamp, author = {Salemi, Alireza and Mysore, Sheshera and Bendersky, Michael and Zamani, Hamed}, title = {{LaMP: When Large Language Models Meet Personalization}}, booktitle = {Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)}, pages = {7370--7392}, publisher = {Association for Computational Linguistics}, year = {2024}, doi = {10.18653/v1/2024.acl-long.399} } @article{zhang2025memorysurvey, author = {Zhang, Zeyu and Dai, Quanyu and Bo, Xiaohe and Ma, Chen and Li, Rui and Chen, Xu and Zhu, Jieming and Dong, Zhenhua and Wen, Ji-Rong}, title = {{A Survey on the Memory Mechanism of Large Language Model-based Agents}}, journal = {ACM Transactions on Information Systems}, volume = {43}, number = {6}, articleno = {155}, numpages = {47}, year = {2025}, doi = {10.1145/3748302} } @incollection{cowan2008memory, author = {Cowan, Nelson}, title = {What are the differences between long-term, short-term, and working memory?}, booktitle = {Essence of Memory}, series = {Progress in Brain Research}, volume = {169}, pages = {323--338}, publisher = {Elsevier}, year = {2008}, doi = {10.1016/S0079-6123(07)00020-9} } @article{baddeley2003workingmemory, author = {Baddeley, Alan}, title = {Working memory: looking back and looking forward}, journal = {Nature Reviews Neuroscience}, volume = {4}, number = {10}, pages = {829--839}, year = {2003}, doi = {10.1038/nrn1201} } @article{squire2015memorysystems, author = {Squire, Larry R. and Dede, Adam J. O.}, title = {Conscious and Unconscious Memory Systems}, journal = {Cold Spring Harbor Perspectives in Biology}, volume = {7}, number = {3}, pages = {a021667}, year = {2015}, doi = {10.1101/cshperspect.a021667} } @inproceedings{bagdasarian2024airgapagent, author = {Bagdasarian, Eugene and Yi, Ren and Ghalebikesabi, Sahra and Kairouz, Peter and Gruteser, Marco and Oh, Sewoong and Balle, Borja and Ramage, Daniel}, title = {{AirGapAgent: Protecting Privacy-Conscious Conversational Agents}}, booktitle = {Proceedings of the 2024 on ACM SIGSAC Conference on Computer and Communications Security}, pages = {3868--3882}, publisher = {ACM}, year = {2024}, doi = {10.1145/3658644.3690350} } @misc{ai2026memorybench, author = {Ai, Q. and Tang, Y. and Wang, C. and Long, J. and Su, W. and Liu, Y.}, title = {{MemoryBench: A Benchmark for Memory and Continual Learning in LLM Systems}}, year = {2026}, howpublished = {Proceedings of the 43rd International Conference on Machine Learning}, url = {https://arxiv.org/abs/2510.17281} } @misc{alqithami2025forgetful, author = {Alqithami, S.}, title = {{Forgetful but Faithful: A Cognitive Memory Architecture and Benchmark for Privacy-Aware Generative Agents}}, year = {2025}, howpublished = {arXiv preprint arXiv:2512.12856}, url = {https://arxiv.org/abs/2512.12856} } @misc{aman2025bitmar, author = {Aman, E. and Carlin, E. and Pao, H.-K. and Beltrame, G. and Sari, G. I. P. and Chen, Y.-T.}, title = {{BitMar: Low-Bit Multimodal Fusion with Episodic Memory for Edge Devices}}, year = {2025}, howpublished = {Proceedings of the BabyLM Workshop at EMNLP 2025}, url = {https://arxiv.org/abs/2510.10560} } @misc{arakawa2024prism, author = {Arakawa, R. and Yakura, H. and Goel, M.}, title = {{PrISM-Observer: Intervention Agent to Help Users Perform Everyday Procedures Sensed using a Smartwatch}}, year = {2024}, howpublished = {Proceedings of the 37th Annual ACM Symposium on User Interface Software and Technology, 1-16}, doi = {10.1145/3654777.3676350} } @misc{banerjee2026apex, author = {Banerjee, P. and Moshtaghi, M. and Subramanian, S. and Misra, A. and Chadha, A.}, title = {{APEX-MEM: Agentic Semi-Structured Memory with Temporal Reasoning for Long-Term Conversational AI}}, year = {2026}, howpublished = {In Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) (pp. 16470--16489)}, doi = {10.18653/v1/2026.acl-long.749} } @misc{bang2024crayon, author = {Bang, J. and Lee, J. and Shim, K. and Yang, S. and Chang, S.}, title = {{Crayon: Customized On-Device LLM via Instant Adapter Blending and Edge-Server Hybrid Inference}}, year = {2024}, howpublished = {In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) (pp. 3720--3731)}, doi = {10.18653/v1/2024.acl-long.204} } @misc{bhardwaj2026superlocalmemory, author = {Bhardwaj, V. P.}, title = {{SuperLocalMemory: Privacy-Preserving Multi-Agent Memory with Bayesian Trust Defense Against Memory Poisoning}}, year = {2026}, howpublished = {arXiv preprint arXiv:2603.02240}, url = {https://arxiv.org/abs/2603.02240} } @misc{bini2025memlora, author = {Bini, M. and Bohdal, O. and Michieli, U. and Akata, Z. and Ozay, M. and Ceritli, T.}, title = {{MemLoRA: Distilling Expert Adapters for On-Device Memory Systems}}, year = {2025}, howpublished = {arXiv preprint arXiv:2512.04763}, url = {https://arxiv.org/abs/2512.04763} } @misc{bohdal2026clustering, author = {Bohdal, O. and Saha, P. and Michieli, U. and Ozay, M. and Ceritli, T.}, title = {{Clustering-driven Memory Compression for On-device Large Language Models}}, year = {2026}, howpublished = {arXiv preprint arXiv:2601.17443}, url = {https://arxiv.org/abs/2601.17443} } @misc{cai2026color, author = {Cai, W. and Zhang, H. and Huang, Y. and Sun, S. and Deng, J. and Xu, S. and Song, J. and Zhang, Z.}, title = {{Color When It Counts: Grayscale-Guided Online Triggering for Always-On Streaming Video Sensing}}, year = {2026}, howpublished = {arXiv preprint arXiv:2603.22466}, url = {https://arxiv.org/abs/2603.22466} } @misc{cakmak2025tripss, author = {Cakmak, M. C. and Agarwal, N. and Poudel, D.}, title = {{TriPSS: A Tri-Modal Keyframe Extraction Framework Using Perceptual, Structural, and Semantic Representations}}, year = {2025}, howpublished = {Proceedings of the MA-LLM Workshop at ACM Multimedia 2025}, url = {https://arxiv.org/abs/2506.05395} } @misc{ceritli2025hydraopt, author = {Ceritli, T. and Bohdal, O. and Ozay, M. and Moon, J. and Lee, K. and Ko, H. and Michieli, U.}, title = {{HydraOpt: Navigating the Efficiency-Performance Trade-off of Adapter Merging}}, year = {2025}, howpublished = {In Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing (pp. 26887--26909)}, doi = {10.18653/v1/2025.emnlp-main.1365} } @misc{chen2026memflow, author = {Chen, J. and Li, Y. and Wang, G.}, title = {{MemFlow: Intent-Driven Memory Orchestration for Small Language Model Agents}}, year = {2026}, howpublished = {arXiv preprint arXiv:2605.03312}, url = {https://arxiv.org/abs/2605.03312} } @misc{chen2026deployment, author = {Chen, L. and Zhang, G. and Zhao, K. and Cirne, D. and Olsen, A. and Chu, X. and Miller, Z. and Blanken, A. and Anoun, A. and Ting, J.}, title = {{Deployment-Time Memorization in Foundation-Model Agents}}, year = {2026}, howpublished = {ICML 2026 Workshop on Memorization in Foundation Models}, url = {https://arxiv.org/abs/2606.10062} } @misc{chen2026memprivacy, author = {Chen, Y. and Zhao, J. and Tang, B. and Wang, H. and Zhang, Y. and Huang, F. and Xiong, F. and Li, Z.}, title = {{MemPrivacy: Privacy-Preserving Personalized Memory Management for Edge-Cloud Agents}}, year = {2026}, howpublished = {arXiv preprint arXiv:2605.09530}, url = {https://arxiv.org/abs/2605.09530} } @misc{chhikara2025mem0, author = {Chhikara, P. and Khant, D. and Aryan, S. and Singh, T. and Yadav, D.}, title = {{Mem0: Building Production-Ready AI Agents with Scalable Long-Term Memory}}, year = {2025}, howpublished = {arXiv preprint arXiv:2504.19413}, url = {https://arxiv.org/abs/2504.19413} } @misc{chung2026ecg, author = {Chung, H. and Oh, J. and Kyung, D. and Kim, J. and Kwon, Y. and Kim, M.-G. and Choi, E.}, title = {{ECG-Agent: On-Device Tool-Calling Agent for ECG Multi-Turn Dialogue}}, year = {2026}, howpublished = {Proceedings of the IEEE International Conference on Acoustics, Speech and Signal Processing}, url = {https://arxiv.org/abs/2601.20323} } @misc{chung2026agent, author = {Chung, J. and Shin, B. and Kim, J. and Rhu, M.}, title = {{Agent-X: Full Pipeline Acceleration of On-device AI Agents}}, year = {2026}, howpublished = {arXiv preprint arXiv:2605.10380}, url = {https://arxiv.org/abs/2605.10380} } @misc{deng2026mobilemem, author = {Deng, X. and Xue, Y. and Chen, Y. and Mao, M. and Zhong, R. and Xu, B. and Fang, J. and Xu, H. and Wu, T. and Xu, Y. and Deng, S. and Wang, H. and Chen, H. and Zhang, N.}, title = {{MobileMem: Evaluating Long-Horizon Memory for Language Agents in Real-World Mobile Environments}}, year = {2026}, howpublished = {Lifelong Agent Workshop at ICLR 2026}, url = {https://openreview.net/forum?id=w5I11HrMgJ} } @misc{dorbala2026memctrl, author = {Dorbala, V. S. and Manocha, D.}, title = {{MemCtrl: Using MLLMs as Active Memory Controllers on Embodied Agents}}, year = {2026}, howpublished = {arXiv preprint arXiv:2601.20831}, url = {https://arxiv.org/abs/2601.20831} } @misc{du2025memguide, author = {Du, Y. and Wang, B. and He, Y. and Liang, B. and Wang, B. and Li, Z. and Gui, L. and Pan, J. Z. and Xu, R. and Wong, K.-F.}, title = {{MemGuide: Intent-Driven Memory Selection for Goal-Oriented Multi-Session LLM Agents}}, year = {2025}, howpublished = {arXiv preprint arXiv:2505.20231}, url = {https://arxiv.org/abs/2505.20231} } @misc{elhaddad2025ar, author = {El Haddad, R. A. and Wang, Z. and Shin, Y. and Liu, R. and Wang, Y. and Yu, C.}, title = {{AR Secretary Agent: Real-time Memory Augmentation via LLM-powered Augmented Reality Glasses}}, year = {2025}, howpublished = {arXiv preprint arXiv:2505.11888}, url = {https://arxiv.org/abs/2505.11888} } @misc{fan2026personalized, author = {Fan, H. and Yuan, Y. and Hsia, C. and Liang, C. and Peng, Z.}, title = {{Toward Personalized Proactive Agents on Smart Glasses: From Explicit Context Policies to Implicit User Differences}}, year = {2026}, howpublished = {Proceedings of the Extended Abstracts of the 2026 CHI Conference on Human Factors in Computing Systems, 1-6}, doi = {10.1145/3772363.3798509} } @misc{fang2026lightmem, author = {Fang, J. and Deng, X. and Xu, H. and Jiang, Z. and Tang, Y. and Xu, Z. and Deng, S. and Yao, Y. and Wang, M. and Qiao, S. and Chen, H. and Zhang, N.}, title = {{LightMem: Lightweight and Efficient Memory-Augmented Generation}}, year = {2026}, howpublished = {International Conference on Learning Representations}, url = {https://arxiv.org/abs/2510.18866} } @misc{gong2024population, author = {Gong, J. and Ding, J. and Meng, F. and Chen, G. and Chen, H. and Zhao, S. and Lu, H. and Li, Y.}, title = {{A Population-to-individual Tuning Framework for Adapting Pretrained LM to On-device User Intent Prediction}}, year = {2024}, howpublished = {Proceedings of the 30th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, 896-907}, doi = {10.1145/3637528.3671984} } @misc{grauman2022ego4d, author = {Grauman, K. and Westbury, A. and Byrne, E. and Chavis, Z. and Furnari, A. and Girdhar, R. and Hamburger, J. and Jiang, H. and Liu, M. and Liu, X. and Martin, M. and Nagarajan, T. and Radosavovic, I. and Ramakrishnan, S. K. and Ryan, F. and Sharma, J. and Wray, M. and Xu, M. and Xu, E. Z. and others}, title = {{Ego4D: Around the World in 3,000 Hours of Egocentric Video}}, year = {2022}, howpublished = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition}, url = {https://arxiv.org/abs/2110.07058} } @misc{gupta2025memo, author = {Gupta, G. and Yadav, K. and Kira, Z. and Gal, Y. and Aljundi, R.}, title = {{Memo: Training memory-efficient embodied agents with reinforcement learning}}, year = {2025}, howpublished = {Advances in Neural Information Processing Systems, 38}, url = {https://proceedings.neurips.cc/paper_files/paper/2025/hash/96889893231d651898b0de42fdbee3a6-Abstract-Conference.html} } @misc{he2026memoryarena, author = {He, Z. and Wang, Y. and Zhi, C. and Hu, Y. and Chen, T.-P. and Yin, L. and Chen, Z. and Wu, T. A. and Ouyang, S. and Wang, Z. and Pei, J. and McAuley, J. and Choi, Y. and Pentland, A.}, title = {{MemoryArena: Benchmarking Agent Memory in Interdependent Multi-Session Agentic Tasks}}, year = {2026}, howpublished = {arXiv preprint arXiv:2602.16313}, url = {https://arxiv.org/abs/2602.16313} } @misc{hu2022lora, author = {Hu, E. J. and Shen, Y. and Wallis, P. and Allen-Zhu, Z. and Li, Y. and Wang, S. and Wang, L. and Chen, W.}, title = {{LoRA: Low-rank adaptation of large language models}}, year = {2022}, howpublished = {International Conference on Learning Representations}, url = {https://openreview.net/forum?id=nZeVKeeFYf9} } @misc{hu2025hiagent, author = {Hu, M. and Chen, T. and Chen, Q. and Mu, Y. and Shao, W. and Luo, P.}, title = {{HiAgent: Hierarchical Working Memory Management for Solving Long-Horizon Agent Tasks with Large Language Model}}, year = {2025}, howpublished = {In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) (pp. 32779--32798)}, doi = {10.18653/v1/2025.acl-long.1575} } @misc{hu2026evaluating, author = {Hu, Yuanzhe and Wang, Y. and McAuley, J.}, title = {{Evaluating memory in LLM agents via incremental multi-turn interactions}}, year = {2026}, howpublished = {International Conference on Learning Representations}, url = {https://openreview.net/forum?id=DT7JyQC3MR} } @misc{hu2026op, author = {Hu, Y. and Long, Z. and Guo, J. and Sui, X. and Fu, X. and Zhao, W. and Zhao, Y. and Qin, B.}, title = {{OP-Bench: Benchmarking Over-Personalization for Memory-Augmented Personalized Conversational Agents}}, year = {2026}, howpublished = {arXiv preprint arXiv:2601.13722}, url = {https://arxiv.org/abs/2601.13722} } @misc{jiang2025apitar, author = {Jiang, H. and Qiu, D. and Stanescu, A. and Wang, Y. and Duh, H. B.-L. and Guan, F.}, title = {{PITAR: an LLM-powered Agent towards Intelligent and Accurate Manipulations in Extended Reality with Multimodal Interactions}}, year = {2025}, howpublished = {SIGGRAPH Asia 2025 XR, 1-2}, doi = {10.1145/3761667.3761944} } @misc{jiang2025bmemory, author = {Jiang, H. and Zhang, X. and Garg, S. and Arora, R. and Kuo, S.-Z. and Xu, J. and Bansal, A. and Brossman, C. and Liu, Y. and Colak, A. and Aly, A. and Kumar, A. and Dong, X. L.}, title = {{Memory-QA: Answering Recall Questions Based on Multimodal Memories}}, year = {2025}, howpublished = {arXiv preprint arXiv:2509.18436}, url = {https://arxiv.org/abs/2509.18436} } @misc{jiang2026intention, author = {Jiang, P. and Liu, F. and Guo, J. and Wen, C.-K. and Jin, S. and Zhang, J.}, title = {{Intention-Aware Semantic Agent Communications for AI Glasses}}, year = {2026}, howpublished = {arXiv preprint arXiv:2604.23691}, url = {https://arxiv.org/abs/2604.23691} } @misc{joachims2007evaluating, author = {Joachims, T. and Granka, L. and Pan, B. and Hembrooke, H. and Radlinski, F. and Gay, G.}, title = {{Evaluating the accuracy of implicit feedback from clicks and query reformulations in Web search}}, year = {2007}, howpublished = {ACM Transactions on Information Systems, 25(2), 7}, doi = {10.1145/1229179.1229181} } @misc{jonelagadda2025mnemosyne, author = {Jonelagadda, A. and Hahn, C. and Zheng, H. and Penachio, S.}, title = {{Mnemosyne: An unsupervised, human-inspired long-term memory architecture for edge-based LLMs}}, year = {2025}, howpublished = {arXiv preprint arXiv:2510.08601}, url = {https://arxiv.org/abs/2510.08601} } @misc{kang2025memory, author = {Kang, J. and Ji, M. and Zhao, Z. and Bai, T.}, title = {{Memory OS of AI Agent}}, year = {2025}, howpublished = {In Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing (pp. 25961--25970)}, doi = {10.18653/v1/2025.emnlp-main.1318} } @misc{kaviani2026opal, author = {Kaviani, D. and Ozdarendeli, A. E. and Zhu, J. and Ding, Y. and Popa, R. A.}, title = {{Opal: Private Memory for Personal AI}}, year = {2026}, howpublished = {arXiv preprint arXiv:2604.02522}, url = {https://arxiv.org/abs/2604.02522} } @misc{kelly2003implicit, author = {Kelly, D. and Teevan, J.}, title = {{Implicit feedback for inferring user preference}}, year = {2003}, howpublished = {ACM SIGIR Forum, 37(2), 18-28}, doi = {10.1145/959258.959260} } @misc{khan2026clickaixr, author = {Khan, D. and Kouyoumdjian, A. and Liu, X. and Mena, O. and Engel, D. and Viola, I.}, title = {{ClickAIXR: On-Device Multimodal Vision-Language Interaction with Real-World Objects in Extended Reality}}, year = {2026}, howpublished = {arXiv preprint arXiv:2604.04905}, url = {https://arxiv.org/abs/2604.04905} } @misc{kim2026speechless, author = {Kim, Y. and Jadeja, D. and Pradhan, D. and Yang, Y. and Kaufman, A. E.}, title = {{SpeechLess: Micro-utterance with Personalized Spatial Memory-aware Assistant in Everyday Augmented Reality}}, year = {2026}, howpublished = {2026 IEEE Conference on Virtual Reality and 3D User Interfaces (VR), 217-227}, doi = {10.1109/VR67842.2026.00044} } @misc{kumar2026memarchitect, author = {Kumar, L. S. and Ba, Y. and Pan, R.}, title = {{MemArchitect: A Policy Driven Memory Governance Layer}}, year = {2026}, howpublished = {arXiv preprint arXiv:2603.18330}, url = {https://arxiv.org/abs/2603.18330} } @misc{kwon2026embodied, author = {Kwon, T. and Choi, D. and Kim, H. and Kim, S. and Moon, S. and Kwak, B.-W. and Huang, K.-H. and Yeo, J.}, title = {{Embodied Agents Meet Personalization: Investigating Challenges and Solutions Through the Lens of Memory Utilization}}, year = {2026}, howpublished = {International Conference on Learning Representations}, url = {https://arxiv.org/abs/2505.16348} } @misc{li2026anchors, author = {Li, W. and Zhang, Z. and Zhang, M. and Hong, G. and Wen, M. and You, X. and Yang, M.}, title = {{From Anchors to Supervision: Memory-Graph Guided Corpus-Free Unlearning for Large Language Models}}, year = {2026}, howpublished = {arXiv preprint arXiv:2604.13777}, url = {https://arxiv.org/abs/2604.13777} } @misc{li2026efficient, author = {Li, X. and Zhang, Y. and Zhou, F. and Zhang, P. and Ji, X.}, title = {{Efficient on-device continuous learning enabling evolutionary edge intelligence}}, year = {2026}, howpublished = {Neurocomputing, Article 132515}, doi = {10.1016/j.neucom.2026.132515} } @misc{li2024appagent, author = {Li, Y. and Zhang, C. and Jiang, W. and Yang, W. and Fu, B. and Cheng, P. and Chen, X. and Chen, L. and Wei, Y.}, title = {{AppAgent v2: Advanced Agent for Flexible Mobile Interactions}}, year = {2024}, howpublished = {arXiv preprint arXiv:2408.11824}, url = {https://arxiv.org/abs/2408.11824} } @misc{li2026ember, author = {Li, Y. and Banerjee, S. and Che, T.}, title = {{EMBER: Efficient Memory via Budgeted Evidence Retention for Long-Horizon Agents}}, year = {2026}, howpublished = {arXiv preprint arXiv:2606.05894}, url = {https://arxiv.org/abs/2606.05894} } @misc{liang2026learning, author = {Liang, K. and Kruk, J. and Qian, S. and Yang, X. and Bi, S. and Yao, Y. and Nie, S. and Zhang, M. and Liu, L. and Fisac, J. F. and Zhou, S. and Hosseini, S.}, title = {{Learning Personalized Agents from Human Feedback}}, year = {2026}, howpublished = {arXiv preprint arXiv:2602.16173}, url = {https://arxiv.org/abs/2602.16173} } @misc{liu2026awhat, author = {Liu, C. and Chen, L. and Zhou, H. and Zhang, X. and Kong, Q. and Tong, P. and Wang, W. and Yu, X. and Hoi, S. and Wang, Y.}, title = {{What Memory Do GUI Agents Really Need? From Passive Records to Active Task-Driving States}}, year = {2026}, howpublished = {arXiv preprint arXiv:2606.31612}, url = {https://arxiv.org/abs/2606.31612} } @misc{liu2026bworldmemarena, author = {Liu, C. and Yang, Y. and Pu, S. X. and Liu, Y. and Long, L. and Guo, Y. and Chen, N. and Weng, Z. and Kochkina, E. and Kaur, S. and Smiley, C. and Liu, X. and Zou, J. and Liu, S. and Bu, Y. and Peng, S. and Wang, X. E.}, title = {{WorldMemArena: Evaluating Multimodal Agent Memory Through Action-World Interaction}}, year = {2026}, howpublished = {arXiv preprint arXiv:2605.29341}, url = {https://arxiv.org/abs/2605.29341} } @misc{liu2026pm, author = {Liu, G. and Gabriel, S.}, title = {{PM-Bench: Evaluating Prospective Memory in LLM Agents}}, year = {2026}, howpublished = {Conference on Language Modeling}, url = {https://arxiv.org/abs/2607.12385} } @misc{liu2026memgui, author = {Liu, G. and Zhao, P. and Liang, Y. and Luo, Q. and Tang, S. and Chai, Y. and Lin, W. and Xiao, H. and Wang, W. and Chen, S. and Lu, Z. and Wu, G. and Wang, H. and Liu, L. and Liu, Y.}, title = {{MemGUI-Bench: Benchmarking Memory of Mobile GUI Agents in Dynamic Environments}}, year = {2026}, howpublished = {arXiv preprint arXiv:2602.06075}, url = {https://arxiv.org/abs/2602.06075} } @misc{liu2026perma, author = {Liu, S. and Zhu, J. and Shu, L. and Lin, J. and Chen, Y. and Zhang, H. and Zhang, C. and Xu, D. and Li, J. and Tang, B. and Li, Z. and Xiong, F. and Chen, E. and Xu, T.}, title = {{PERMA: Benchmarking Personalized Memory Agents via Event-Driven Preference and Realistic Task Environments}}, year = {2026}, howpublished = {arXiv preprint arXiv:2603.23231}, url = {https://arxiv.org/abs/2603.23231} } @misc{liu2025beyond, author = {Liu, Z. and Zhang, C. and Zhao, X. and Feng, Y. and Bai, B. and Feng, D. and Feng, E. and Xia, Y. and Chen, H.}, title = {{Beyond Training: Enabling Self-Evolution of Agents with MOBIMEM}}, year = {2025}, howpublished = {arXiv preprint arXiv:2512.15784}, url = {https://arxiv.org/abs/2512.15784} } @misc{long2025seeing, author = {Long, L. and He, Y. and Ye, W. and Pan, Y. and Lin, Y. and Li, H. and Zhao, J. and Li, W.}, title = {{Seeing, Listening, Remembering, and Reasoning: A Multimodal Agent with Long-Term Memory}}, year = {2025}, howpublished = {arXiv preprint arXiv:2508.09736}, url = {https://arxiv.org/abs/2508.09736} } @misc{lyu2026personalalign, author = {Lyu, Y. and Chen, G. and Shao, R. and Guan, W. and Nie, L.}, title = {{PersonalAlign: Hierarchical Implicit Intent Alignment for Personalized GUI Agent with Long-Term User-Centric Records}}, year = {2026}, howpublished = {In Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) (pp. 36074--36089)}, doi = {10.18653/v1/2026.acl-long.1669} } @misc{maharana2024evaluating, author = {Maharana, A. and Lee, D.-H. and Tulyakov, S. and Bansal, M. and Barbieri, F. and Fang, Y.}, title = {{Evaluating Very Long-Term Conversational Memory of LLM Agents}}, year = {2024}, howpublished = {In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) (pp. 13851--13870)}, doi = {10.18653/v1/2024.acl-long.747} } @misc{maheri2026zk, author = {Maheri, M. M. and Cotterill, S. and Davidson, A. and Haddadi, H.}, title = {{ZK-APEX: Zero-Knowledge Approximate Personalized Unlearning with Executable Proofs}}, year = {2026}, howpublished = {Proceedings of the 9th Conference on Machine Learning and Systems}, url = {https://arxiv.org/abs/2512.09953} } @misc{mei2026according, author = {Mei, J. and Chen, J. and Yang, G. and Hou, X. and Li, M. and Byrne, B.}, title = {{According to Me: Long-Term Personalized Referential Memory QA}}, year = {2026}, howpublished = {arXiv preprint arXiv:2603.01990}, url = {https://arxiv.org/abs/2603.01990} } @misc{merrill2026transforming, author = {Merrill, M. A. and Paruchuri, A. and Rezaei, N. and Kovacs, G. and Perez, J. and Liu, Y. and Schenck, E. and Hammerquist, N. and Sunshine, J. and Tailor, S. and Ayush, K. and Su, H.-W. and He, Q. and McLean, C. Y. and Malhotra, M. and Patel, S. and Zhan, J. and Althoff, T. and McDuff, D. and Liu, X.}, title = {{Transforming wearable data into personal health insights using large language model agents}}, year = {2026}, howpublished = {Nature Communications, 17(1), 1143}, doi = {10.1038/s41467-025-67922-y} } @misc{mohammed2025context, author = {Mohammed, H. and Yin, H. and Boyapati, S. C.}, title = {{Context adaptive memory-efficient LLM inference for edge multi-agent systems: Extended abstract}}, year = {2025}, howpublished = {In Proceedings of the 24th International Conference on Autonomous Agents and Multiagent Systems (pp. 2678--2680)}, url = {https://dl.acm.org/doi/10.5555/3709347.3743976} } @misc{mollyn2022samosa, author = {Mollyn, V. and Ahuja, K. and Verma, D. and Harrison, C. and Goel, M.}, title = {{SAMoSA: Sensing activities with motion and subsampled audio}}, year = {2022}, howpublished = {Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies, 6(3), 1-19}, doi = {10.1145/3550284} } @misc{nie2026pspa, author = {Nie, H. and Liu, X. and Bai, Y. and Wang, Y. and Liu, Y. and Yao, Q. and Wang, Z.}, title = {{PSPA-Bench: A Personalized Benchmark for Smartphone GUI Agent}}, year = {2026}, howpublished = {arXiv preprint arXiv:2603.29318}, url = {https://arxiv.org/abs/2603.29318} } @misc{park2025mobilerag, author = {Park, T. and Lee, G. and Kim, M.-S.}, title = {{MobileRAG: A Fast, Memory-Efficient, and Energy-Efficient Method for On-Device RAG}}, year = {2025}, howpublished = {arXiv preprint arXiv:2507.01079}, url = {https://arxiv.org/abs/2507.01079} } @misc{paruchuri2025egotrigger, author = {Paruchuri, A. and Hersek, S. and Aggarwal, L. and Yang, Q. and Liu, X. and Kulshrestha, A. and Colaco, A. and Fuchs, H. and Chatterjee, I.}, title = {{EgoTrigger: Toward Audio-Driven Image Capture for Human Memory Enhancement in All-Day Energy-Efficient Smart Glasses}}, year = {2025}, howpublished = {Proceedings of the IEEE International Symposium on Mixed and Augmented Reality}, url = {https://arxiv.org/abs/2508.01915} } @misc{pu2025promemassist, author = {Pu, K. and Zhang, T. and Sendhilnathan, N. and Freitag, S. and Sodhi, R. and Jonker, T. R.}, title = {{ProMemAssist: Exploring Timely Proactive Assistance Through Working Memory Modeling in Multi-Modal Wearable Devices}}, year = {2025}, howpublished = {Proceedings of the 38th Annual ACM Symposium on User Interface Software and Technology, 1-19}, doi = {10.1145/3746059.3747770} } @misc{rachapudi2026repair, author = {Rachapudi, J. and Singh, P. and Vatsi, R. and Hambarde, P. and Shukla, A.}, title = {{RePAIR: Interactive Machine Unlearning through Prompt-Aware Model Repair}}, year = {2026}, howpublished = {arXiv preprint arXiv:2604.12820}, url = {https://arxiv.org/abs/2604.12820} } @misc{rasmussen2025zep, author = {Rasmussen, P. and Paliychuk, P. and Beauvais, T. and Ryan, J. and Chalef, D.}, title = {{Zep: A Temporal Knowledge Graph Architecture for Agent Memory}}, year = {2025}, howpublished = {arXiv preprint arXiv:2501.13956}, url = {https://arxiv.org/abs/2501.13956} } @misc{rawles2025androidworld, author = {Rawles, C. and Clinckemaillie, S. and Chang, Y. and Waltz, J. and Lau, G. and Fair, M. and Li, A. and Bishop, W. and Li, W. and Campbell-Ajala, F. and Toyama, D. and Berry, R. and Tyamagundlu, D. and Lillicrap, T. and Riva, O.}, title = {{AndroidWorld: A Dynamic Benchmarking Environment for Autonomous Agents}}, year = {2025}, howpublished = {International Conference on Learning Representations}, url = {https://arxiv.org/abs/2405.14573} } @misc{rege2026agentic, author = {Rege, A. and Sadhu, A. and Li, Y. and Li, K. and Vinayak, R. K. and Chai, Y. and Lee, Y. J. and Kim, H. J.}, title = {{Agentic Very Long Video Understanding}}, year = {2026}, howpublished = {arXiv preprint arXiv:2601.18157}, url = {https://arxiv.org/abs/2601.18157} } @misc{saadfalcon2026openjarvis, author = {Saad-Falcon, J. and Narayan, A. and Manihani, R. and Bhathal, T. and Shandilya, H. and Akengin, H. O. and Bo, G. and Park, A. and Hart, M. and Costello, C. and Li, C. and Ré, C. and Mirhoseini, A.}, title = {{OpenJarvis: Personal AI, On Personal Devices}}, year = {2026}, howpublished = {arXiv preprint arXiv:2605.17172}, url = {https://arxiv.org/abs/2605.17172} } @misc{shen2026mem2actbench, author = {Shen, Y. and Li, K. and Zhou, W. and Hu, S.}, title = {{Mem2ActBench: A Benchmark for Evaluating Long-Term Memory Utilization in Task-Oriented Autonomous Agents}}, year = {2026}, howpublished = {In Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) (pp. 8173--8190)}, doi = {10.18653/v1/2026.acl-long.370} } @misc{shenaj2026k, author = {Shenaj, D. and Bohdal, O. and Ceritli, T. and Ozay, M. and Zanuttigh, P. and Michieli, U.}, title = {{K-Merge: Online Continual Merging of Adapters for On-device Large Language Models}}, year = {2026}, howpublished = {In Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) (pp. 3013--3029)}, doi = {10.18653/v1/2026.acl-long.137} } @misc{shi2026vlons, author = {Shi, J. and Zhang, H. and Zhang, Y. and Lam, T. L. and Zhang, L. and Huang, H. and Gao, Y.}, title = {{VLONS: A Vision-and-Language On-Device Navigation System With Multimodal Fusion and Modular Framework}}, year = {2026}, howpublished = {IEEE Transactions on Consumer Electronics, 72(1), 1171-1182}, doi = {10.1109/TCE.2025.3638139} } @misc{shkolnikov2026agent, author = {Shkolnikov, Y. P.}, title = {{Agent Memory Below the Prompt: Persistent Q4 KV Cache for Multi-Agent LLM Inference on Edge Devices}}, year = {2026}, howpublished = {arXiv preprint arXiv:2603.04428}, url = {https://arxiv.org/abs/2603.04428} } @misc{shu2026remem, author = {Shu, Y. and Jonnalagedda, S. P. and Gao, X. and Gutiérrez, B. J. and Qi, W. and Das, K. and Sun, H. and Su, Y.}, title = {{REMem: Reasoning with Episodic Memory in Language Agent}}, year = {2026}, howpublished = {International Conference on Learning Representations}, url = {https://arxiv.org/abs/2602.13530} } @misc{simsek2026when, author = {Simsek, B.}, title = {{When to Forget: A Memory Governance Primitive}}, year = {2026}, howpublished = {arXiv preprint arXiv:2604.12007}, url = {https://arxiv.org/abs/2604.12007} } @misc{song2026bitlora, author = {Song, I. and Lee, K.}, title = {{BitLoRA: Quantization-Compatible Adapter Tuning for 1.58-bit LLM in Federated On-Device AI-Agent}}, year = {2026}, howpublished = {Expert Systems with Applications, 311, 131397}, doi = {10.1016/j.eswa.2026.131397} } @misc{sun2026magnet, author = {Sun, L. and Zhang, J. and Wang, S. and Wei, Z.}, title = {{MAGNET: Towards Adaptive GUI Agents with Memory-Driven Knowledge Evolution}}, year = {2026}, howpublished = {arXiv preprint arXiv:2601.19199}, url = {https://arxiv.org/abs/2601.19199} } @misc{sun2026egograph, author = {Sun, S. and Han, K. and Huang, Y. and Cai, W. and Song, J.}, title = {{EgoGraph: Temporal Knowledge Graph for Egocentric Video Understanding}}, year = {2026}, howpublished = {arXiv preprint arXiv:2602.23709}, url = {https://arxiv.org/abs/2602.23709} } @misc{tan2025membench, author = {Tan, H. and Zhang, Z. and Ma, C. and Chen, X. and Dai, Q. and Dong, Z.}, title = {{MemBench: Towards More Comprehensive Evaluation on the Memory of LLM-based Agents}}, year = {2025}, howpublished = {In Findings of the Association for Computational Linguistics: ACL 2025 (pp. 19336--19352)}, doi = {10.18653/v1/2025.findings-acl.989} } @misc{tan2025prospect, author = {Tan, Z. and Yan, J. and Hsu, I.-H. and Han, R. and Wang, Z. and Le, L. and Song, Y. and Chen, Y. and Palangi, H. and Lee, G. and Iyer, A. R. and Chen, T. and Liu, H. and Lee, C.-Y. and Pfister, T.}, title = {{In Prospect and Retrospect: Reflective Memory Management for Long-term Personalized Dialogue Agents}}, year = {2025}, howpublished = {In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) (pp. 8416--8439)}, doi = {10.18653/v1/2025.acl-long.413} } @misc{tang2026egomemory, author = {Tang, Y. and Zhang, J. and Qin, X. and Yu, J. and Qiu, M. and Gou, G. and Xiong, G. and Lin, Q. and Rajmohan, S. and Zhang, D. and Wu, Q.}, title = {{EgoMemory: Memory-Augmented Personalized Retrieval for Long-Context Egocentric Video}}, year = {2026}, howpublished = {In Findings of the Association for Computational Linguistics: ACL 2026 (pp. 7317--7349)}, doi = {10.18653/v1/2026.findings-acl.362} } @misc{tiwari2025quantspec, author = {Tiwari, R. and Xi, H. and Tomar, A. and Hooper, C. and Kim, S. and Horton, M. and Najibi, M. and Mahoney, M. W. and Keutzer, K. and Gholami, A.}, title = {{QuantSpec: Self-Speculative Decoding with Hierarchical Quantized KV Cache}}, year = {2025}, howpublished = {arXiv preprint arXiv:2502.10424}, url = {https://arxiv.org/abs/2502.10424} } @misc{uddin2026recall, author = {Uddin, M. N. and Shubham, K. and Blanco, E. and Baral, C. and Wang, G.}, title = {{From Recall to Forgetting: Benchmarking Long-Term Memory for Personalized Agents}}, year = {2026}, howpublished = {In Findings of the Association for Computational Linguistics: ACL 2026 (pp. 26814--26841)}, doi = {10.18653/v1/2026.findings-acl.1337} } @misc{vijayvargiya2025efficient, author = {Vijayvargiya, S. and Lokesh, R.}, title = {{Efficient On-Device Agents via Adaptive Context Management}}, year = {2025}, howpublished = {arXiv preprint arXiv:2511.03728}, url = {https://arxiv.org/abs/2511.03728} } @article{wang2023voyager, author = {Wang, G. and Xie, Y. and Jiang, Y. and Mandlekar, A. and Xiao, C. and Zhu, Y. and Fan, L. and Anandkumar, A.}, title = {{Voyager: An Open-Ended Embodied Agent with Large Language Models}}, year = {2024}, journal = {Transactions on Machine Learning Research}, url = {https://openreview.net/forum?id=ehfRiF0R3a} } @misc{wang2024mobile, author = {Wang, J. and Xu, H. and Jia, H. and Zhang, X. and Yan, M. and Shen, W. and Zhang, J. and Huang, F. and Sang, J.}, title = {{Mobile-Agent-v2: Mobile Device Operation Assistant with Effective Navigation via Multi-Agent Collaboration}}, year = {2024}, howpublished = {arXiv preprint arXiv:2406.01014}, url = {https://arxiv.org/abs/2406.01014} } @misc{wang2026stamp, author = {Wang, J. and Xu, H. and Zhang, X. and Zhu, Z. and Yan, M. and Ye, J. and Sang, J.}, title = {{STAMP: Training Explicit Memory for Mobile GUI Agents in Controllable and Scalable Virtual Environments}}, year = {2026}, howpublished = {arXiv preprint arXiv:2605.29324}, url = {https://arxiv.org/abs/2605.29324} } @misc{wang2026me, author = {Wang, S. and Liu, C. and Loo, G. and Zheng, L. and Wei, K. and Zeng, X. and Zhang, J. and Tian, Y.}, title = {{Me-Agent: A Personalized Mobile Agent with Two-Level User Habit Learning for Enhanced Interaction}}, year = {2026}, howpublished = {arXiv preprint arXiv:2601.20162}, url = {https://arxiv.org/abs/2601.20162} } @misc{wang2025dynakv, author = {Wang, T. and Huang, M. and Li, F. and Chen, L. and Zhang, J. and Ren, J.}, title = {{DynaKV: Enabling Accurate and Efficient Long-Sequence LLM Decoding on Smartphones}}, year = {2025}, howpublished = {arXiv preprint arXiv:2511.07427}, url = {https://arxiv.org/abs/2511.07427} } @misc{wang2025perpilot, author = {Wang, X. and Cui, Z. and Li, H. and Zeng, Y. and Wang, C. and Song, R. and Chen, Y. and Shao, K. and Zhang, Q. and Liu, J. and Ren, S. and Hu, S. and Wang, Z.}, title = {{PerPilot: Personalizing VLM-based Mobile Agents via Memory and Exploration}}, year = {2025}, howpublished = {arXiv preprint arXiv:2508.18040}, url = {https://arxiv.org/abs/2508.18040} } @misc{wang2026s, author = {Wang, X. and Zhao, X. and Rodriguez, P. and Sachan, D. S. and Oguz, B. and Moon, S. and Li, S.-W. and Ghosh, G. and Dong, X. and Yih, W.-T.}, title = {{S-EMBER: A Large-Scale Benchmark for Streaming Egocentric Memory Retrieval}}, year = {2026}, howpublished = {arXiv preprint arXiv:2607.02689}, url = {https://arxiv.org/abs/2607.02689} } @misc{wang2026egoself, author = {Wang, Y. and Xu, Y. and Li, X. and Hong, J. and Wang, Y. and Chen, C. W. and Zhu, W.}, title = {{EgoSelf: From Memory to Personalized Egocentric Assistant}}, year = {2026}, howpublished = {arXiv preprint arXiv:2604.19564}, url = {https://arxiv.org/abs/2604.19564} } @misc{wang2025leann, author = {Wang, Y. and Li, Z. and Liu, S. and Wu, Y. and Mao, Z. and Zhao, Y. and Yan, X. and Xu, Z. and Zhou, Y. and Stoica, I. and Min, S. and Zaharia, M. and Gonzalez, J. E.}, title = {{LEANN: A Low-Storage Vector Index}}, year = {2025}, howpublished = {arXiv preprint arXiv:2506.08276}, url = {https://arxiv.org/abs/2506.08276} } @misc{wang2025mirix, author = {Wang, Y. and Chen, X.}, title = {{MIRIX: Multi-Agent Memory System for LLM-Based Agents}}, year = {2025}, howpublished = {arXiv preprint arXiv:2507.07957}, url = {https://arxiv.org/abs/2507.07957} } @misc{wang2024crafting, author = {Wang, Z. and Li, Z. and Jiang, Z. and Tu, D. and Shi, W.}, title = {{Crafting Personalized Agents through Retrieval-Augmented Generation on Editable Memory Graphs}}, year = {2024}, howpublished = {In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing (pp. 4891--4906)}, doi = {10.18653/v1/2024.emnlp-main.281} } @misc{wu2026forget, author = {Wu, B. and Ding, Z. and Huang, J. and Zhao, Y.}, title = {{Forget to Improve: On-Device LLM-Agent Continual Learning via Budget-Curated Memory}}, year = {2026}, howpublished = {arXiv preprint arXiv:2606.25115}, url = {https://arxiv.org/abs/2606.25115} } @misc{wu2025longmemeval, author = {Wu, D. and Wang, H. and Yu, W. and Zhang, Y. and Chang, K.-W. and Yu, D.}, title = {{LongMemEval: Benchmarking Chat Assistants on Long-Term Interactive Memory}}, year = {2025}, howpublished = {International Conference on Learning Representations}, url = {https://arxiv.org/abs/2410.10813} } @misc{wu2026longmemeval, author = {Wu, D. and Ji, Z. and Kawatkar, A. and Kwan, B. and Gu, J.-C. and Peng, N. and Chang, K.-W.}, title = {{LongMemEval-V2: Evaluating Long-Term Agent Memory Toward Experienced Colleagues}}, year = {2026}, howpublished = {arXiv preprint arXiv:2605.12493}, url = {https://arxiv.org/abs/2605.12493} } @misc{wu2026agent, author = {Wu, J. and Hu, M. and Zhu, J. and Wang, J. and Jin, Y.}, title = {{Agent-Memory Protocol: A privacy-focused protocol for LLM agents and user memory interaction}}, year = {2026}, howpublished = {In Proceedings of the Second AAAI Bridge Program on AI for Medicine and Healthcare (PMLR Vol. 317)}, url = {https://proceedings.mlr.press/v317/wu26a.html} } @misc{wu2025auto, author = {Wu, W. and Zhou, K. and Yuan, R. and Yu, V. and Wang, S. and Hu, Z. and Huang, B.}, title = {{Auto-scaling Continuous Memory for GUI Agent}}, year = {2025}, howpublished = {arXiv preprint arXiv:2510.09038}, url = {https://arxiv.org/abs/2510.09038} } @misc{wu2025quick, author = {Wu, Z. and Huang, H. and Yang, Y. and Song, Y. and Lou, X. and Liu, W. and Zhang, W. and Wang, J. and Zhang, Z.}, title = {{Quick on the Uptake: Eliciting Implicit Intents from Human Demonstrations for Personalized Mobile-Use Agents}}, year = {2025}, howpublished = {arXiv preprint arXiv:2508.08645}, url = {https://arxiv.org/abs/2508.08645} } @misc{xia2024edge, author = {Xia, X. and Wang, Z. and Sun, R. and Liu, B. and Khalil, I. and Xue, M.}, title = {{Edge Unlearning is Not "on Edge"! An Adaptive Exact Unlearning System on Resource-Constrained Devices}}, year = {2024}, howpublished = {arXiv preprint arXiv:2410.10128}, url = {https://arxiv.org/abs/2410.10128} } @misc{xiao2026ui, author = {Xiao, H. and Wang, G. and Wang, H. and Liu, S. and Chai, Y. and Pan, Y. and Zhou, Y. and Chen, X. and Wen, Y. and Li, H.}, title = {{UI-Mem: Self-Evolving Experience Memory for Online Reinforcement Learning in Mobile GUI Agents}}, year = {2026}, howpublished = {arXiv preprint arXiv:2602.05832}, url = {https://arxiv.org/abs/2602.05832} } @misc{xu2025mem, author = {Xu, W. and Liang, Z. and Mei, K. and Gao, H. and Tan, J. and Zhang, Y.}, title = {{A-Mem: Agentic memory for LLM agents}}, year = {2025}, howpublished = {Advances in Neural Information Processing Systems, 38}, url = {https://proceedings.neurips.cc/paper_files/paper/2025/hash/19909c36f51abc4856b4560aff3d36d6-Abstract-Conference.html} } @misc{yan2026memory, author = {Yan, S. and Yang, X. and Huang, Z. and Nie, E. and Ding, Z. and Li, Z. and Ma, X. and Bi, J. and Kersting, K. and Pan, J. Z. and Schuetze, H. and Tresp, V. and Ma, Y.}, title = {{Memory-R1: Enhancing Large Language Model Agents to Manage and Utilize Memories via Reinforcement Learning}}, year = {2026}, howpublished = {In Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) (pp. 12805--12825)}, doi = {10.18653/v1/2026.acl-long.583} } @misc{yang2025contextagent, author = {Yang, B. and Xu, L. and Zeng, L. and Liu, K. and Jiang, S. and Lu, W. and Chen, H. and Jiang, X. and Xing, G. and Yan, Z.}, title = {{ContextAgent: Context-Aware Proactive LLM Agents with Open-World Sensory Perceptions}}, year = {2025}, howpublished = {Advances in Neural Information Processing Systems, 38}, url = {https://arxiv.org/abs/2505.14668} } @misc{yang2026proagent, author = {Yang, B. and Xu, L. and Zeng, L. and Guo, Y. and Jiang, S. and Lu, W. and Liu, K. and Li, Y. and Jiang, X. and Xing, G. and Yan, Z.}, title = {{ProAgent: Harnessing On-Demand Sensory Contexts for Proactive LLM Agent Systems in the Wild}}, year = {2026}, howpublished = {arXiv preprint arXiv:2512.06721}, url = {https://arxiv.org/abs/2512.06721} } @misc{yang2025egolife, author = {Yang, J. and Liu, S. and Guo, H. and Dong, Y. and Zhang, X. and Zhang, S. and Wang, P. and Zhou, Z. and Xie, B. and Wang, Z. and Ouyang, B. and Lin, Z. and Cominelli, M. and Cai, Z. and Zhang, Y. and Zhang, P. and Hong, F. and Widmer, J. and Gringoli, F. and others}, title = {{EgoLife: Towards Egocentric Life Assistant}}, year = {2025}, howpublished = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 28885--28900}, url = {https://arxiv.org/abs/2503.03803} } @misc{yang2026selfmem, author = {Yang, S. and Wu, J. and Wong, D. F. and Wang, D.}, title = {{SelfMem: Self-Optimizing Memory for AI Agents}}, year = {2026}, howpublished = {arXiv preprint arXiv:2607.03726}, url = {https://arxiv.org/abs/2607.03726} } @misc{yang2025efficientnav, author = {Yang, Z. and Zheng, S. and Xie, T. and Xu, T. and Yu, B. and Wang, F. and Tang, J. and Liu, S. and Li, M.}, title = {{EfficientNav: Towards on-device object-goal navigation with navigation map caching and retrieval}}, year = {2025}, howpublished = {Advances in Neural Information Processing Systems, 38}, url = {https://proceedings.neurips.cc/paper_files/paper/2025/hash/067437c6d5d0369b6d09200bef89715b-Abstract-Conference.html} } @misc{ye2026secure, author = {Ye, D. and Zhu, T. and Zhu, C. and He, F. and He, Q. and Wang, S. and Liu, B. and Zhou, W.}, title = {{Secure Forgetting: A Framework for Privacy-Driven Unlearning in Large Language Model (LLM)-Based Agents}}, year = {2026}, howpublished = {arXiv preprint arXiv:2604.00430}, url = {https://arxiv.org/abs/2604.00430} } @misc{ye2026venus, author = {Ye, S. and Ouyang, B. and Qian, T. and Zeng, L. and Yuan, M. and Chu, X. and Hong, W. and Chen, X.}, title = {{Venus: An Efficient Edge Memory-and-Retrieval System for VLM-based Online Video Understanding}}, year = {2026}, howpublished = {IEEE INFOCOM 2026}, url = {https://arxiv.org/abs/2512.07344} } @misc{yin2026focal, author = {Yin, H. and Wen, Z. and Cao, J. and Yuan, B. and Yang, R.}, title = {{FOCAL: Filtered On-device Continuous Activity Logging for Efficient Personal Desktop Summarization}}, year = {2026}, howpublished = {arXiv preprint arXiv:2604.19541}, url = {https://arxiv.org/abs/2604.19541} } @misc{yu2026agentic, author = {Yu, Y. and Yao, L. and Xie, Y. and Tan, Q. and Feng, J. and Li, Y. and Wu, L.}, title = {{Agentic Memory: Learning Unified Long-Term and Short-Term Memory Management for Large Language Model Agents}}, year = {2026}, howpublished = {In Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) (pp. 21457--21483)}, doi = {10.18653/v1/2026.acl-long.981} } @misc{yuan2026star, author = {Yuan, M. and Zhang, H. and Mohammadi, M. and Li, R. and Shan, J. and Waslander, S. L.}, title = {{STaR: Scalable Task-Conditioned Retrieval for Long-Horizon Multi-Modal Robot Memory}}, year = {2026}, howpublished = {IEEE Robotics and Automation Letters, 11(5), 5994-6001}, doi = {10.1109/LRA.2026.3677723} } @misc{zhang2023appagent, author = {Zhang, C. and Yang, Z. and Liu, J. and Li, Y. and Han, Y. and Chen, X. and Huang, Z. and Fu, B. and Yu, G.}, title = {{AppAgent: Multimodal Agents as Smartphone Users}}, year = {2023}, howpublished = {arXiv preprint arXiv:2312.13771}, url = {https://arxiv.org/abs/2312.13771} } @misc{zhang2025memory, author = {Zhang, C. and Yang, C. and Tan, Q. and Liu, J. and Li, A. and Wang, Y. and Lu, J. and Wang, J. and Yuan, G.}, title = {{Towards Memory-Efficient and Sustainable Machine Unlearning on Edge using Zeroth-Order Optimizer}}, year = {2025}, howpublished = {Proceedings of the Great Lakes Symposium on VLSI 2025, 227-232}, doi = {10.1145/3716368.3735273} } @misc{zhang2025kvswap, author = {Zhang, H. and Xia, C. and Wang, Z.}, title = {{KVSwap: Disk-aware KV Cache Offloading for Long-Context On-device Inference}}, year = {2025}, howpublished = {arXiv preprint arXiv:2511.11907}, url = {https://arxiv.org/abs/2511.11907} } @misc{zhang2026forget, author = {Zhang, T. and Tong, Y. and Dong, J. and Xu, K. and Wang, Y. and Yuan, J.}, title = {{Forget by Uncertainty: Orthogonal Entropy Unlearning for Quantized Neural Networks}}, year = {2026}, howpublished = {Proceedings of the 43rd International Conference on Machine Learning}, url = {https://arxiv.org/abs/2602.00567} } @misc{zhao2026museheterogeneityawaremultimediasearch, title={MUSE: A Heterogeneity-Aware Multimedia Search Engine for Mobile SoCs}, author={Xinkui Zhao and Qingyu Ma and Yifan Zhang and Hengxuan Lou and Sai Liu and Chang Liu and Guanjie Cheng and Naibo Wang and Yueshen Xu}, year={2026}, eprint={2511.19192}, archivePrefix={arXiv}, primaryClass={cs.DC}, url={https://arxiv.org/abs/2511.19192}, } @misc{zhao2026amaskclaw, author = {Zhao, Y. and Zheng, D. and Huang, K. and Wei, Y. and Yang, Z. and Zhou, L.}, title = {{MaskClaw: Edge-Side Personalized Privacy Arbitration for GUI Agents with Behavior-Driven Skill Evolution}}, year = {2026}, howpublished = {arXiv preprint arXiv:2605.28646}, url = {https://arxiv.org/abs/2605.28646} } @misc{zhao2026bama, author = {Zhao, Y. and Yuan, B. and Huang, J. and Yuan, H. and Yu, Z. and Xu, H. and Hu, L. and Shankarampeta, A. and Huang, Z. and Ni, W. and Tian, Y. and Zhao, J.}, title = {{AMA-Bench: Evaluating Long-Horizon Memory for Agentic Applications}}, year = {2026}, howpublished = {arXiv preprint arXiv:2602.22769}, url = {https://arxiv.org/abs/2602.22769} } @misc{zheng2026evaluating, author = {Zheng, J. and Liu, Z. and Shen, Z. and Qu, J. and Chen, G. and Wang, Y. and Xu, Y. and Liu, Y. and Cheng, S.}, title = {{Evaluating Memory Capability in Continuous Lifelog Scenario}}, year = {2026}, howpublished = {In Findings of the Association for Computational Linguistics: ACL 2026 (pp. 7063--7089)}, doi = {10.18653/v1/2026.findings-acl.351} } @misc{zheng2026task, author = {Zheng, Y. and Liu, Z. and Zhao, X. and Wen, J.-R.}, title = {{A Task-State Representation for Long-Horizon Mobile GUI Agents}}, year = {2026}, howpublished = {arXiv preprint arXiv:2607.00502}, url = {https://arxiv.org/abs/2607.00502} } @misc{zhong2026actionengine, author = {Zhong, H. and Faisal, F. and França, L. and Leesatapornwongsa, T. and Szekeres, A. and Rong, K. and Nath, S.}, title = {{ActionEngine: From Reactive to Programmatic GUI Agents via State Machine Memory}}, year = {2026}, howpublished = {arXiv preprint arXiv:2602.20502}, url = {https://arxiv.org/abs/2602.20502} } @misc{zhong2026device, author = {Zhong, Z. and Yuan, W. and Qu, L. and Chen, T. and Wang, H. and Zhao, X. and Yin, H.}, title = {{Towards On-device Personalization: Cloud-device Collaborative Data Augmentation for Efficient On-device Language Model}}, year = {2026}, howpublished = {ACM Transactions on Intelligent Systems and Technology, 17(1), 1-22}, doi = {10.1145/3779452} } @misc{ziliotto2025personal, author = {Ziliotto, F. and Akkara, J. R. and Daniele, A. and Ballan, L. and Serafini, L. and Campari, T.}, title = {{PersONAL: Towards a Comprehensive Benchmark for Personalized Embodied Agents}}, year = {2025}, howpublished = {arXiv preprint arXiv:2509.19843}, url = {https://arxiv.org/abs/2509.19843} } %% Added for the revised foundations, 2026-09-23. @article{sumers2024cognitive, author = {Sumers, Theodore R. and Yao, Shunyu and Narasimhan, Karthik and Griffiths, Thomas L.}, title = {{Cognitive Architectures for Language Agents}}, journal = {Transactions on Machine Learning Research}, year = {2024}, url = {https://arxiv.org/abs/2309.02427} } @article{wang2024autonomoussurvey, author = {Wang, Lei and Ma, Chen and Feng, Xueyang and Zhang, Zeyu and Yang, Hao and Zhang, Jingsen and Chen, Zhiyuan and Tang, Jiakai and Chen, Xu and Lin, Yankai and Zhao, Wayne Xin and Wei, Zhewei and Wen, Jirong}, title = {{A survey on large language model based autonomous agents}}, journal = {Frontiers of Computer Science}, volume = {18}, number = {6}, articleno = {186345}, year = {2024}, doi = {10.1007/s11704-024-40231-1} } @article{xi2025agentsurvey, author = {Xi, Zhiheng and Chen, Wenxiang and Guo, Xin and He, Wei and Ding, Yiwen and Hong, Boyang and Zhang, Ming and Wang, Junzhe and Jin, Senjie and Zhou, Enyu and Zheng, Rui and Fan, Xiaoran and Wang, Xiao and Xiong, Limao and Zhou, Yuhao and Wang, Weiran and Jiang, Changhao and Zou, Yicheng and Liu, Xiangyang and Yin, Zhangyue and Dou, Shihan and Weng, Rongxiang and Qin, Wenjuan and Zheng, Yongyan and Qiu, Xipeng and Huang, Xuanjing and Zhang, Qi and Gui, Tao}, title = {{The rise and potential of large language model based agents: a survey}}, journal = {Science China Information Sciences}, volume = {68}, number = {2}, articleno = {121101}, year = {2025}, doi = {10.1007/s11432-024-4222-0} } @inproceedings{park2023generativeagents, author = {Park, Joon Sung and O'Brien, Joseph and Cai, Carrie Jun and Morris, Meredith Ringel and Liang, Percy and Bernstein, Michael S.}, title = {{Generative Agents: Interactive Simulacra of Human Behavior}}, booktitle = {Proceedings of the 36th Annual ACM Symposium on User Interface Software and Technology}, publisher = {Association for Computing Machinery}, address = {New York, NY, USA}, pages = {1--22}, year = {2023}, url = {https://arxiv.org/abs/2304.03442} } @inproceedings{wang2025awm, author = {Wang, Zora Zhiruo and Mao, Jiayuan and Fried, Daniel and Neubig, Graham}, title = {{Agent Workflow Memory}}, booktitle = {Proceedings of the 42nd International Conference on Machine Learning}, series = {Proceedings of Machine Learning Research}, volume = {267}, pages = {63897--63911}, publisher = {PMLR}, year = {2025}, url = {https://proceedings.mlr.press/v267/wang25bx.html} } @article{zhao2024expel, author = {Zhao, Andrew and Huang, Daniel and Xu, Quentin and Lin, Matthieu and Liu, Yong-Jin and Huang, Gao}, title = {{ExpeL: LLM Agents Are Experiential Learners}}, journal = {Proceedings of the AAAI Conference on Artificial Intelligence}, volume = {38}, number = {17}, pages = {19632--19642}, year = {2024}, doi = {10.1609/aaai.v38i17.29936} } @misc{zhou2025simplestrongbaselinelongterm, title={A Simple Yet Strong Baseline for Long-Term Conversational Memory of LLM Agents}, author={Sizhe Zhou and Jiawei Han}, year={2025}, eprint={2511.17208}, archivePrefix={arXiv}, primaryClass={cs.CL}, url={https://arxiv.org/abs/2511.17208}, } @misc{zhang2026memarenaegocentricbenchmarkondevice, title={MemArena: An Ego-Centric Benchmark for On-Device Agentic Personal Memory Assistants at Scale}, author={Jiadong Zhang and Xiaosong Ma}, year={2026}, eprint={2608.02613}, archivePrefix={arXiv}, primaryClass={cs.CL}, url={https://arxiv.org/abs/2608.02613}, } @misc{hu2024hiagenthierarchicalworkingmemory, title={HiAgent: Hierarchical Working Memory Management for Solving Long-Horizon Agent Tasks with Large Language Model}, author={Mengkang Hu and Tianxing Chen and Qiguang Chen and Yao Mu and Wenqi Shao and Ping Luo}, year={2024}, eprint={2408.09559}, archivePrefix={arXiv}, primaryClass={cs.CL}, url={https://arxiv.org/abs/2408.09559}, } @misc{xu2026structmemstructuredmemorylonghorizon, title={StructMem: Structured Memory for Long-Horizon Behavior in LLMs}, author={Buqiang Xu and Yijun Chen and Jizhan Fang and Ruobin Zhong and Yunzhi Yao and Yuqi Zhu and Lun Du and Shumin Deng}, year={2026}, eprint={2604.21748}, archivePrefix={arXiv}, primaryClass={cs.CL}, url={https://arxiv.org/abs/2604.21748}, } @misc{li2025camconstructivistviewagentic, title={CAM: A Constructivist View of Agentic Memory for LLM-Based Reading Comprehension}, author={Rui Li and Zeyu Zhang and Xiaohe Bo and Zihang Tian and Xu Chen and Quanyu Dai and Zhenhua Dong and Ruiming Tang}, year={2025}, eprint={2510.05520}, archivePrefix={arXiv}, primaryClass={cs.CL}, url={https://arxiv.org/abs/2510.05520}, } @misc{cao2026higmemhierarchicalllmguidedmemory, title={HiGMem: A Hierarchical and LLM-Guided Memory System for Long-Term Conversational Agents}, author={Shuqi Cao and Jingyi He and Fei Tan}, year={2026}, eprint={2604.18349}, archivePrefix={arXiv}, primaryClass={cs.CL}, url={https://arxiv.org/abs/2604.18349}, } @misc{ahn2025hemahippocampusinspiredextended, title={HEMA : A Hippocampus-Inspired Extended Memory Architecture for Long-Context AI Conversations}, author={Kwangseob Ahn}, year={2025}, eprint={2504.16754}, archivePrefix={arXiv}, primaryClass={cs.CL}, url={https://arxiv.org/abs/2504.16754}, } @inproceedings{wang2026lightwm, title={LightWM: Training-Free Hierarchical Working Memory for Small Language Model Agents}, author={Wang, Ziyi and Jin, Haonan and Wang, Zian and Wang, Wendong and Zhang, Lanshan}, booktitle={Proceedings of the 43rd International Conference on Machine Learning}, year={2026}, organization={PMLR}, note={ICML 2026} } @misc{shi2025personaxrecommendationagentoriented, title={PersonaX: A Recommendation Agent Oriented User Modeling Framework for Long Behavior Sequence}, author={Yunxiao Shi and Wujiang Xu and Zeqi Zhang and Xing Zi and Qiang Wu and Min Xu}, year={2025}, eprint={2503.02398}, archivePrefix={arXiv}, primaryClass={cs.IR}, url={https://arxiv.org/abs/2503.02398}, } @misc{mao2026bimembidirectionalconstructionhierarchical, title={Bi-Mem: Bidirectional Construction of Hierarchical Memory for Personalized LLMs via Inductive-Reflective Agents}, author={Wenyu Mao and Haosong Tan and Shuchang Liu and Haoyang Liu and Yifan Xu and Huaxiang Ji and Xiang Wang}, year={2026}, eprint={2601.06490}, archivePrefix={arXiv}, primaryClass={cs.MA}, url={https://arxiv.org/abs/2601.06490}, } @misc{wang2026qumempersonalizedmemoryqueryconditioned, title={QUMem: Personalized Memory for Query-Conditioned User-State Inference in LLM Agents}, author={Heng Wang and Yifei Li and Lingling Zhang and Pengyu Li and Xinyu Che and Xinyu Zhang and Zesheng Yang}, year={2026}, eprint={2608.16168}, archivePrefix={arXiv}, primaryClass={cs.CL}, url={https://arxiv.org/abs/2608.16168}, } @misc{wang2025omemomnimemorypersonalized, title={O-Mem: Omni Memory System for Personalized, Long Horizon, Self-Evolving Agents}, author={Piaohong Wang and Motong Tian and Jiaxian Li and Yuan Liang and Yuqing Wang and Qianben Chen and Tiannan Wang and Zhicong Lu and Jiawei Ma and Yuchen Eleanor Jiang and Wangchunshu Zhou}, year={2025}, eprint={2511.13593}, archivePrefix={arXiv}, primaryClass={cs.CL}, url={https://arxiv.org/abs/2511.13593}, } @misc{lee2026volumevaluepreferencealignedmemory, title={From Volume to Value: Preference-Aligned Memory Construction for On-Device RAG}, author={Changmin Lee and Jaemin Kim and Taesik Gong}, year={2026}, eprint={2605.18271}, archivePrefix={arXiv}, primaryClass={cs.CL}, url={https://arxiv.org/abs/2605.18271}, } @misc{chen2026corememriemannianretrievalfisherguided, title={CoreMem: Riemannian Retrieval and Fisher-Guided Distillation for Long-Term Memory in Dialogue Agents}, author={Jiaqi Chen and Yongqin Zeng and Shaoshen Chen and Yijian Zhang and Hai-Tao Zheng and Chunxia Ma and XiuTeng Zhou}, year={2026}, eprint={2606.18406}, archivePrefix={arXiv}, primaryClass={cs.CL}, url={https://arxiv.org/abs/2606.18406}, } @misc{10.5281/zenodo.19686729, doi = {10.5281/ZENODO.19686729}, url = {https://zenodo.org/doi/10.5281/zenodo.19686729}, author = {Patel, Ishan and Joshi, Ishan}, keywords = {vector quantization, KV cache compression, multi-agent LLM, TurboQuant, large language models, transformer inference, perplexity, asymmetric quantization, FWHT, shared memory}, title = {PolyKV: Shared Asymmetric TurboQuant-Compressed KV Memory for Multi-Agent LLMs}, publisher = {Ishan Patel}, year = {2026}, copyright = {MIT License} } @misc{boroujeni2026dontwastebitsadaptive, title={Don't Waste Bits! Adaptive KV-Cache Quantization for Lightweight On-Device LLMs}, author={Sayed Pedram Haeri Boroujeni and Niloufar Mehrabi and Patrick Woods and Gabriel Hillesheim and Abolfazl Razi}, year={2026}, eprint={2604.04722}, archivePrefix={arXiv}, primaryClass={cs.CV}, url={https://arxiv.org/abs/2604.04722}, } @misc{lee2025ciflexcontextualinstructionflow, title={CIFLEX: Contextual Instruction Flow for Sub-task Execution in Multi-Turn Interactions with a Single On-Device LLM}, author={Juntae Lee and Jihwan Bang and Seunghan Yang and Simyung Chang}, year={2025}, eprint={2510.01239}, archivePrefix={arXiv}, primaryClass={cs.CL}, url={https://arxiv.org/abs/2510.01239}, } @misc{pan2025kvflowefficientprefixcaching, title={KVFlow: Efficient Prefix Caching for Accelerating LLM-Based Multi-Agent Workflows}, author={Zaifeng Pan and Ajjkumar Patel and Zhengding Hu and Yipeng Shen and Yue Guan and Wan-Lu Li and Lianhui Qin and Yida Wang and Yufei Ding}, year={2025}, eprint={2507.07400}, archivePrefix={arXiv}, primaryClass={cs.DC}, url={https://arxiv.org/abs/2507.07400}, } @misc{ye2025kvcommonlinecrosscontextkvcache, title={KVCOMM: Online Cross-context KV-cache Communication for Efficient LLM-based Multi-agent Systems}, author={Hancheng Ye and Zhengqi Gao and Mingyuan Ma and Qinsi Wang and Yuzhe Fu and Ming-Yu Chung and Yueqian Lin and Zhijian Liu and Jianyi Zhang and Danyang Zhuo and Yiran Chen}, year={2025}, eprint={2510.12872}, archivePrefix={arXiv}, primaryClass={cs.MA}, url={https://arxiv.org/abs/2510.12872}, } @misc{aliev2026recachelearningbudgetawarecaching, title={ReCache: Learning Budget-Aware Caching Schedules for Diffusion Models via REINFORCE}, author={Mishan Aliev and Eva Neudachina and Ilya Bykov and Aleksandr Oganov and Kirill Struminsky and Aibek Alanov and Denis Rakitin}, year={2026}, eprint={2606.06060}, archivePrefix={arXiv}, primaryClass={cs.CV}, url={https://arxiv.org/abs/2606.06060}, } @misc{zhang2026sparsexefficientsegmentlevelkv, title={SparseX: Efficient Segment-Level KV Cache Sharing for Interleaved LLM Serving}, author={Quqing Zhang and Kai Chen and Ning Liao and Zehao Lin and Bo Tang and Feiyu Xiong and Zhiyu Li and Xiaoxing Wang}, year={2026}, eprint={2606.01751}, archivePrefix={arXiv}, primaryClass={cs.PF}, url={https://arxiv.org/abs/2606.01751}, } @misc{shu2026tierkvlongcontextondevicellms, title={TierKV: Long-Context On-Device LLMs via Predictive Multi-Tier KV Caching}, author={Zhihao Shu and Md Musfiqur Rahman Sanim and Jie Hu and Kun Yuan and Minghai Qin and Gagan Agrawal and Wei Niu}, year={2026}, eprint={2609.21172}, archivePrefix={arXiv}, primaryClass={cs.LG}, url={https://arxiv.org/abs/2609.21172}, } @misc{ren2026scalingselfevolvingagentsparametric, title={Scaling Self-Evolving Agents via Parametric Memory}, author={Tao Ren and Weiyao Luo and Hui Yang and Rongzhi Zhu and Xiang Huang and Yuchuan Wu and Bingxue Chou and Jieping Ye and Jiafeng Liang and Yongbin Li and Yijie Peng}, year={2026}, eprint={2606.04536}, archivePrefix={arXiv}, primaryClass={cs.AI}, url={https://arxiv.org/abs/2606.04536}, } @misc{guo2026peamparametricembodiedagent, title={PEAM: Parametric Embodied Agent Memory through Contrastive Internalization of Experience in Minecraft}, author={Yuchen Guo and Junli Gong and Weicheng Wang and Hongmin Cai and Yiu-ming Cheung and Weifeng Su}, year={2026}, eprint={2605.27762}, archivePrefix={arXiv}, primaryClass={cs.AI}, url={https://arxiv.org/abs/2605.27762}, } @misc{hosseini2026duomemcapableondevicememory, title={DuoMem: Towards Capable On-Device Memory Agents via Dual-Space Distillation}, author={Peyman Hosseini and Ondrej Bohdal and Ahmed Alajrami and Andrea Maracani and Ignacio Castro and Matthew Purver and Mete Ozay and Savas Ozkan and Taha Ceritli}, year={2026}, eprint={2606.29961}, archivePrefix={arXiv}, primaryClass={cs.LG}, url={https://arxiv.org/abs/2606.29961}, } @misc{zhao2026hymemhybridmemoryarchitecture, title={HyMem: Hybrid Memory Architecture with Dynamic Retrieval Scheduling}, author={Xiaochen Zhao and Kaikai Wang and Xiaowen Zhang and Chen Yao and Aili Wang}, year={2026}, eprint={2602.13933}, archivePrefix={arXiv}, primaryClass={cs.AI}, url={https://arxiv.org/abs/2602.13933}, } @misc{fang2026recache, title = {{ReCache: Efficient KV Cache Reuse and Compression for Tool-Augmented LLM Agents}}, author = {Fang, Yichu and Wei, Sitong and Hu, Haozhe and Shen, Xiaoyu}, year = {2026}, eprint = {2608.19662}, archivePrefix = {arXiv}, primaryClass = {cs.CL}, url = {https://arxiv.org/abs/2608.19662} } % Core literature integration, 2026-09-27 % S368 @misc{pubgally2026ally, author = {{PUBG Ally Team}}, title = {{PUBG Ally: A Conversational Embodied Agent as an AI Teammate}}, year = {2026}, howpublished = {arXiv preprint arXiv:2609.29837}, eprint = {2609.29837}, archivePrefix = {arXiv}, url = {https://arxiv.org/abs/2609.29837}, } % S102 @inproceedings{kim2025infinipotv, author = {Kim, Minsoo and Shim, Kyuhong and Choi, Jungwook and Chang, Simyung}, title = {{InfiniPot-V: Memory-Constrained KV Cache Compression for Streaming Video Understanding}}, year = {2025}, booktitle = {Advances in Neural Information Processing Systems}, volume = {38}, url = {https://neurips.cc/virtual/2025/poster/116667}, } % S153 @inproceedings{kim2026epicache, author = {Kim, Minsoo and Kundu, Arnav and Kim, Han-Byul and Dixit, Richa and Cho, Minsik}, title = {{EpiCache: Episodic KV Cache Management for Long-Term Conversation on Resource-Constrained Environments}}, year = {2026}, booktitle = {International Conference on Machine Learning}, url = {https://icml.cc/virtual/2026/poster/65405}, } % S175 @misc{liu2026local, author = {Liu, Xinxin and Li, Jiaxin and Wang, Zibo and Ji, Yun and Zhu, Zhangqi and Hu, Qing and Wang, Zhibin and Gu, Rong and Zhong, Sheng and Tian, Chen}, title = {{LOCAL: Enabling Learning On-device Contiguously for Agent LLMs}}, year = {2026}, howpublished = {arXiv preprint arXiv:2608.15241}, eprint = {2608.15241}, archivePrefix = {arXiv}, url = {https://arxiv.org/abs/2608.15241}, } % S070 @inproceedings{qin2026eam, author = {Qin, Zerui and Yue, Sheng and Hua, Xingyuan and Fu, Yongjian and Ren, Ju}, title = {{Executable Agentic Memory for GUI Agent}}, year = {2026}, booktitle = {International Conference on Machine Learning}, url = {https://icml.cc/virtual/2026/poster/65931}, } % S047 @inproceedings{ahn2026mobilekgqa, author = {Ahn, Junyong and Han, Hyeongrok and Kang, Bong Gyun and Mok, Jisoo and Lee, Byunghan and Yoon, Sungroh}, title = {{MobileKGQA: On-Device KGQA System on Dynamic Mobile Environments}}, year = {2026}, booktitle = {International Conference on Learning Representations}, url = {https://proceedings.iclr.cc/paper_files/paper/2026/hash/24e4e3234178a836b70e0aa48827e0ff-Abstract-Conference.html}, } % S046 @inproceedings{lu2026mobiedit, author = {Lu, Zhenyan and Xu, Daliang and Cai, Dongqi and Li, Zexi and Liu, Wei and Luan, Jian and Liu, Fangming and Wang, Shangguang and Xu, Mengwei}, title = {{MobiEdit: Resource-efficient Knowledge Editing for Personalized On-device LLMs}}, year = {2026}, booktitle = {International Conference on Learning Representations}, url = {https://proceedings.iclr.cc/paper_files/paper/2026/hash/2b35d186908df6fa14a51ba1cae9dc4e-Abstract-Conference.html}, } % S485 @misc{shan2026huoziime, author = {Shan, Baocai and Xu, Yuzhuang and Che, Wanxiang}, title = {{HUOZIIME: An On-Device LLM-enhanced Input Method for Deep Personalization}}, year = {2026}, howpublished = {arXiv preprint arXiv:2604.14159}, eprint = {2604.14159}, archivePrefix = {arXiv}, url = {https://arxiv.org/abs/2604.14159}, } % S299 @article{liu2026memaura, author = {Liu, Siyuan and Chen, Huangxun}, title = {{MemAura: Structured Context Memory for Personalized LLM Reasoning in Smart Environments}}, year = {2026}, journal = {Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies}, volume = {10}, number = {2}, articleno = {55}, numpages = {43}, doi = {10.1145/3810196}, url = {https://doi.org/10.1145/3810196}, } % S342 @inproceedings{nishikawa2026hmaps, author = {Nishikawa, Koji and Kato, Makoto P.}, title = {{H-MAPS: Hierarchical Memory-Augmented Proactive Search Assistant for Scientific Literature}}, year = {2026}, booktitle = {Proceedings of the 49th International ACM SIGIR Conference on Research and Development in Information Retrieval}, doi = {10.1145/3805712.3808378}, url = {https://doi.org/10.1145/3805712.3808378}, } % S333 @inproceedings{xu2026orisa, author = {Xu, Haichuan and Oygenblik, David and Zhang, Runze and Yao, Mingxuan and Ibrahim, Muhammad and Saltaformaggio, Brendan}, title = {{Recovering and Rehosting Mobile Local LLM Conversations and Contexts via Memory Forensics}}, year = {2026}, booktitle = {2026 IEEE Symposium on Security and Privacy}, doi = {10.1109/SP63933.2026.00252}, url = {https://doi.org/10.1109/SP63933.2026.00252}, } % S385 @misc{bhardwaj2026superlocalmemory4, author = {Bhardwaj, Varun Pratap and Singh, Garima and Bhardwaj, Arun Pratap}, title = {{SuperLocalMemory 4.0: The Governed Memory Operating System for AI Agents}}, year = {2026}, howpublished = {arXiv preprint arXiv:2608.08253}, eprint = {2608.08253}, archivePrefix = {arXiv}, url = {https://arxiv.org/abs/2608.08253}, } % S470 @misc{darwin2026mimemory, author = {{Darwin Agent Team}}, title = {{Mi-Memory: A Lifecycle Memory Framework for Personal AI}}, year = {2026}, howpublished = {arXiv preprint arXiv:2607.18975}, eprint = {2607.18975}, archivePrefix = {arXiv}, url = {https://arxiv.org/abs/2607.18975}, } % S375 @article{mun2026affectiveagent, author = {Mun, Reina and Wan, Zishen and Reddi, Vijay Janapa}, title = {{Affective Agent: On-Device Personalized Intervention Reasoning for Wearable Systems}}, year = {2026}, journal = {IEEE Internet Computing}, doi = {10.1109/MIC.2026.3732091}, url = {https://doi.org/10.1109/MIC.2026.3732091}, } % S443 @inproceedings{maniar2025mempal, author = {Maniar, Natasha and Chan, Samantha and Zulfikar, Wazeer and Ren, Scott and Xu, Christine and Maes, Pattie}, title = {{MemPal: Leveraging Multimodal AI and LLMs for Voice-Activated Object Retrieval in Homes of Older Adults}}, year = {2025}, booktitle = {Proceedings of the 30th International Conference on Intelligent User Interfaces}, doi = {10.1145/3708359.3712151}, url = {https://doi.org/10.1145/3708359.3712151}, } % S401 @misc{zhu2026pabench, author = {Zhu, Lichen and Lin, Yueqian and Wang, Yiheng and Liu, Yudong and Li, Hai and Chen, Yiran}, title = {{Will My Assistant Remember My Allergy? What Personal LLM Assistants Forget When Conversation Memory Is Compressed}}, year = {2026}, howpublished = {arXiv preprint arXiv:2609.05767}, eprint = {2609.05767}, archivePrefix = {arXiv}, url = {https://arxiv.org/abs/2609.05767}, } % S373 @misc{deng2026mobilememyear, author = {Deng, Xinle and Xue, Yida and Ru, Xiangyuan and Chen, Yijun and Xu, Buqiang and Mao, Mingjun and Liu, Xinjie and Xu, Haoming and Qiao, Shuofei and Wang, Mengru and Jiang, Chen and Jiang, Yuchen Eleanor and Wang, Lizhong and Wang, Jason and Zeng, Li and Wang, Haofen and Qi, Guilin and Chen, Huajun and Zhang, Ningyu}, title = {{MobileMem: Learning from a Year of Mobile Experiences}}, year = {2026}, howpublished = {arXiv preprint arXiv:2608.13606}, eprint = {2608.13606}, archivePrefix = {arXiv}, url = {https://arxiv.org/abs/2608.13606}, } % Grouped supplementary literature integrated on 2026-09-28 % S420 @misc{loo2025mobileragagent, author = {Loo, Gowen and Liu, Chang and Yin, Qinghong and Chen, Xiang and Chen, Jiawei and Zhang, Jingyuan and Tian, Yu}, title = {{MobileRAG: Enhancing Mobile Agent with Retrieval-Augmented Generation}}, year = {2025}, howpublished = {arXiv preprint arXiv:2509.03891}, eprint = {2509.03891}, archivePrefix = {arXiv}, url = {https://arxiv.org/abs/2509.03891}, } % S387 @misc{malo2026projectmem, author = {Malo, Ripon Chandra and Qiu, Tong}, title = {{PROJECTMEM: A Local-First, Event-Sourced Memory and Judgment Layer for AI Coding Agents}}, year = {2026}, howpublished = {arXiv preprint arXiv:2606.12329}, eprint = {2606.12329}, archivePrefix = {arXiv}, url = {https://arxiv.org/abs/2606.12329}, } % S345 @inproceedings{deng2026lever, numpages = {12}, author = {Deng, Yongheng and Jiang, Tianyuan and Ma, Zhenya and Wu, Hao and Fu, Yongjian and Pan, Hao and Yue, Sheng and Ren, Ju}, title = {{Accelerating Graph-Based RAG Retrieval via Locality-Aware Device-Cloud Collaboration}}, year = {2026}, booktitle = {Proceedings of the 32nd ACM SIGKDD Conference on Knowledge Discovery and Data Mining V.2}, publisher = {Association for Computing Machinery}, doi = {10.1145/3770855.3817674}, url = {https://doi.org/10.1145/3770855.3817674}, } % S356 @inproceedings{liu2026wab, numpages = {4}, author = {Liu, Mugeng and Li, Shuoqi and Yang, Qi and Ma, Yun}, title = {{WAB: Overcoming Memory, Network, and Security Walls in Native Agentic Browsers with WebAssembly}}, year = {2026}, booktitle = {Companion Proceedings of the ACM Web Conference 2026}, publisher = {Association for Computing Machinery}, doi = {10.1145/3774905.3795093}, url = {https://doi.org/10.1145/3774905.3795093}, } % S379 @misc{arslan2026nexus, author = {Arslan, Mustafa}, title = {{Nexus: Depth-Adaptive KV-Cache Splicing and Retrieval-Decoupled Tool Routing for Agentic LLMs on Unified Memory}}, year = {2026}, howpublished = {arXiv preprint arXiv:2608.20397}, eprint = {2608.20397}, archivePrefix = {arXiv}, url = {https://arxiv.org/abs/2608.20397}, } % S486 @misc{conte2026ruva, author = {Conte, Gabriele and Mattiace, Alessio and Carmosino, Gianni and Aghilar, Potito and Servedio, Giovanni and Musicco, Francesco and Anelli, Vito Walter and Di Noia, Tommaso and Donini, Francesco Maria}, title = {{RUVA: Personalized Transparent On-Device Graph Reasoning}}, year = {2026}, howpublished = {arXiv preprint arXiv:2602.15553}, eprint = {2602.15553}, archivePrefix = {arXiv}, url = {https://arxiv.org/abs/2602.15553}, } % S346 @inproceedings{chen2026ecotune, numpages = {11}, author = {Chen, Gong and Lin, Mingkai and Hong, Xiaobin and Li, Wenzhong and Lu, Sanglu}, title = {{EcoTune: Edge-Cloud Collaborative Model Adaptation for Budget-Constrained On-Device SLM Personalization}}, year = {2026}, booktitle = {Proceedings of the ACM Web Conference 2026}, publisher = {Association for Computing Machinery}, doi = {10.1145/3774904.3792264}, url = {https://doi.org/10.1145/3774904.3792264}, } % S334 @article{wang2026pkr, author = {Wang, Huili and Huang, Yuanhong and Hu, Zhiyang and Li, Qing and Fan, Pingyi and Huang, Yongfeng and Wang, Shangguang and Qi, Tao}, title = {{Efficient and Effective Personalized In-Context Learning for On-Device Large Model Services}}, year = {2026}, journal = {IEEE Transactions on Services Computing}, volume = {19}, number = {4}, pages = {2788--2800}, doi = {10.1109/TSC.2026.3697569}, url = {https://doi.org/10.1109/TSC.2026.3697569}, } % S469 @misc{kirmayr2025carmem, author = {Kirmayr, Johannes and Stappen, Lukas and Schneider, Phillip and Matthes, Florian and Andr{\'e}, Elisabeth}, title = {{CarMem: Enhancing Long-Term Memory in LLM Voice Assistants through Category-Bounding}}, year = {2025}, howpublished = {arXiv preprint arXiv:2501.09645}, eprint = {2501.09645}, archivePrefix = {arXiv}, url = {https://arxiv.org/abs/2501.09645}, } % S123 @inproceedings{zhang2026lightweightmemory, author = {Jiaquan Zhang and Chaoning Zhang and Shuxu Chen and Zhenzhen Huang and Pengcheng Zheng and Zhicheng Wang and Ping Guo and Fan Mo and Sung-Ho Bae and Jie Zou and Jiwei Wei and Yang Yang}, title = {{Lightweight LLM Agent Memory with Small Language Models}}, year = {2026}, booktitle = {Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)}, publisher = {Association for Computational Linguistics}, pages = {12914--12929}, doi = {10.18653/v1/2026.acl-long.588}, url = {https://aclanthology.org/2026.acl-long.588/}, } % S376 @misc{sun2026memx, author = {Sun, Lizheng}, title = {{MemX: A Local-First Long-Term Memory System for AI Assistants}}, year = {2026}, howpublished = {arXiv preprint arXiv:2603.16171}, eprint = {2603.16171}, archivePrefix = {arXiv}, url = {https://arxiv.org/abs/2603.16171}, } % S177 @misc{chang2026scrapmem, author = {Chang, Jiale and Ren, Yuxiang}, title = {{ScrapMem: A Bio-inspired Framework for On-device Personalized Agent Memory via Optical Forgetting}}, year = {2026}, howpublished = {arXiv preprint arXiv:2605.03804}, eprint = {2605.03804}, archivePrefix = {arXiv}, url = {https://arxiv.org/abs/2605.03804}, } % S429 @misc{bhardwaj2026superlocalmemory33, author = {Bhardwaj, Varun Pratap}, title = {{SuperLocalMemory V3.3: The Living Brain -- Biologically-Inspired Forgetting, Cognitive Quantization, and Multi-Channel Retrieval for Zero-LLM Agent Memory Systems}}, year = {2026}, howpublished = {arXiv preprint arXiv:2604.04514}, eprint = {2604.04514}, archivePrefix = {arXiv}, url = {https://arxiv.org/abs/2604.04514}, } % S360 @inproceedings{vekaria2025bighelp, author = {Vekaria, Yash and Canino, Aurelio Loris and Levitsky, Jonathan and Ciechonski, Alex and Callejo, Patricia and Mandalari, Anna Maria and Shafiq, Zubair}, title = {{Big Help or Big Brother? Auditing Tracking, Profiling, and Personalization in Generative AI Assistants}}, year = {2025}, booktitle = {34th USENIX Security Symposium (USENIX Security 25)}, publisher = {USENIX Association}, pages = {8115--8134}, url = {https://www.usenix.org/conference/usenixsecurity25/presentation/vekaria}, } % S432 @misc{braas2025fixedpersona, author = {Braas, Martin and Esterle, Lukas}, title = {{Fixed-Persona SLMs with Modular Memory: Scalable NPC Dialogue on Consumer Hardware}}, year = {2025}, howpublished = {arXiv preprint arXiv:2511.10277}, eprint = {2511.10277}, archivePrefix = {arXiv}, url = {https://arxiv.org/abs/2511.10277}, } % S374 @misc{rao2026budgetbench, author = {Rao, Aditya Karnam Gururaj and Jaggi, Arjun}, title = {{BudgetBench: A Budget-Tiered Protocol and Pilot Harness for Memory Strategy Evaluation in Local Large Language Model Agents}}, year = {2026}, howpublished = {arXiv preprint arXiv:2609.13149}, eprint = {2609.13149}, archivePrefix = {arXiv}, url = {https://arxiv.org/abs/2609.13149}, } % S323 @article{dai2026jmsra, author = {Dai, Lipeng and Xiang, Luping and Yang, Kun}, title = {{A token/KV-cache communication media selection and resource allocation strategy for multi-agent collaboration}}, year = {2026}, journal = {Science China Information Sciences}, doi = {10.1007/s11432-026-5057-2}, url = {https://doi.org/10.1007/s11432-026-5057-2}, } % S386 @misc{kulce2026echo, author = {K{\"u}l{\c c}e, Abdulkadir and Esen, Alihan and Fikir, {\c C}a{\u g}la and Kurt, Berke and Arar, Kuzey and Ercan, G{\"o}khan and Tek, Faik Boray}, title = {{ECHO: A Locally-Deployable Agentic Health Assistant with Temporal Memory, Safety Guardrails, and Speech Assessment}}, year = {2026}, howpublished = {arXiv preprint arXiv:2608.06110}, eprint = {2608.06110}, archivePrefix = {arXiv}, url = {https://arxiv.org/abs/2608.06110}, } % S427 @misc{stocker2026persistentcase, author = {Stocker, Felix}, title = {{Towards Persistent Case-Based Memory for Autonomous Data Science: A CBR-Augmented R\&D-Agent with a Locally Deployable Small Language Model}}, year = {2026}, howpublished = {arXiv preprint arXiv:2606.05250}, eprint = {2606.05250}, archivePrefix = {arXiv}, url = {https://arxiv.org/abs/2606.05250}, } % S176 @misc{cai2026palmclaw, author = {Cai, Hongru and Li, Yongqi and Wei, Ran and Li, Wenjie}, title = {{PalmClaw: A Native On-Device Agent Framework for Mobile Phones}}, year = {2026}, howpublished = {arXiv preprint arXiv:2607.13027}, eprint = {2607.13027}, archivePrefix = {arXiv}, url = {https://arxiv.org/abs/2607.13027}, } % S445 @misc{saleh2025usercentrix, author = {Saleh, Alaa and Tarkoma, Sasu and Donta, Praveen Kumar and Lindgren, Anders and Motlagh, Naser Hossein and Dustdar, Schahram and Pirttikangas, Susanna and Lov{\'e}n, Lauri}, title = {{UserCentrix: An Agentic Memory-augmented AI Framework for Smart Spaces}}, year = {2025}, howpublished = {arXiv preprint arXiv:2505.00472}, eprint = {2505.00472}, archivePrefix = {arXiv}, url = {https://arxiv.org/abs/2505.00472}, } % S389 @misc{chen2026wastingasset, author = {Chen, Josef Liyanjun}, title = {{Memory as a Wasting Asset: Pricing Flash Endurance for Embodied Agents, and the Limits of Doing So}}, year = {2026}, howpublished = {arXiv preprint arXiv:2606.18144}, eprint = {2606.18144}, archivePrefix = {arXiv}, url = {https://arxiv.org/abs/2606.18144}, } % S465 @misc{gopal2026structuredmemory, author = {Gopal, Anusha Madan and Pirbadian, Aras and Carlson, Kristofor D. and Lewis, M Anthony and Tapson, Jonathan}, title = {{Structured Memory for Edge Language Models: Persistent Context and Corpus Retrieval via O(1) SSM State Injection}}, year = {2026}, howpublished = {arXiv preprint arXiv:2608.02560}, eprint = {2608.02560}, archivePrefix = {arXiv}, url = {https://arxiv.org/abs/2608.02560}, }