KEEP: A KV-Cache-Centric Memory Management System for Efficient Embodied Planning
Institute for Artificial Intelligence（人工智能研究院）；School of Integrated Circuits, Peking University（集成电路学院，北京大学）；Shenzhen Institute of Artificial Intelligence（深圳人工智能研究院）；Microsoft Research（微软研究院）；Beijing Advanced Innovation Center for Integrated Circuits（北京集成电路先进创新中心）
Zebin Yang, Tong Xie, Baotong Lu, Shaoshan Liu, Bo Yu, Meng Li
Memory-augmented Large Language Models (LLMs) have demonstrated remarkable capability for complex and long-horizon embodied planning. By keeping track of past experiences and environmental states, memory enables LLMs to maintain a global view, thereby avoiding repetitive exploration. However, existing approaches often store the memory as raw text, leading to excessively long prompts and high prefill latency. While it is possible to store and reuse the KV caches, the efficiency benefits are greatly undermined due to frequent KV cache updates. In this paper, we propose KEEP, a KV-cache-centric memory management system for efficient embodied planning. KEEP features 3 key innovations: (1) a Static-Dynamic Memory Construction algorithm that reduces KV cache recomputation by mixed-granularity memory group; (2) a Multi-hop Memory Re-computation algorithm that dynamically identifies important cross-attention among different memory groups and reconstructs memory interactions iteratively; (3) a Layer-balanced Memory Loading that eliminates unbalanced KV cache loading and cross-attention computation across different layers. Extensive experimental results have demonstrated that KEEP achieves 2.68x speedup with negligible accuracy loss compared with text-based memory methods on ALFRED dataset. Compared with the KV re-computation method CacheBlend (EuroSys'25), KEEP shows 4.13% success rate improvement and 1.90x time-to-first-token (TTFT) reduction. Our code is available on https://github.com/PKU-SEC-Lab/KEEP_Embodied_Memory.
TCPO: Thought-Centric Preference Optimization for Effective Embodied Decision-making
Tsinghua University（清华大学）；Meituan（美团）；Northeastern University（东北大学）；Beijing University of Posts and Telecommunications（北京邮电大学）；Meta AI；Wuhan University（武汉大学）
Beijing Key Laboratory of Intelligent Information Technology, School of Computer Science and Technology, Beijing Institute of Technology（北京智能信息科技重点实验室，计算机科学与技术学院，北京理工大学）；Guangdong Laboratory of Machine Perception and Intelligent Computing, Shenzhen MSU-BIT University（广东机器感知与智能计算实验室，深圳MSU-BIT大学）
From Perception to Planning: Evolving Ego-Centric Task-Oriented Spatiotemporal Reasoning via Curriculum Learning
Zhejiang University（浙江大学）；Tianjin University（天津大学）；Qingdao University（青岛大学）；Shanghai Jiao Tong University（上海交通大学）；National University of Singapore（新加坡国立大学）
HybridKV: Hybrid KV Cache Compression for Efficient Multimodal Large Language Model Inference
The State Key Laboratory of Blockchain and Data Security, Zhejiang University（浙江大学区块链与数据安全国家重点实验室）；Hangzhou High-Tech Zone (Binjiang) Institute of Blockchain and Data Security（杭州高新区（滨江）区块链与数据安全研究院）；Hangzhou Dianzi University（杭州电子科技大学）
Memory Centric Power Allocation for Multi-Agent Embodied Question Answering
Department of Electrical and Computer Engineering, The University of Hong Kong, Hong Kong（香港大学电子与计算机工程系）；Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China（中国科学院深圳先进技术研究所）；Southern University of Science and Technology, Shenzhen, China（南方科技大学）；University of Macau, Macau, China（澳门大学）；Istanbul Medipol University, Istanbul, Turkey（伊斯坦布尔Medipol大学）
OxyGen: Unified KV Cache Management for VLA Inference under Multi-Task Parallelism
Institute for AI Industry Research (AIR)（人工智能产业研究院）；Department of Electronic Engineering（电子工程系）；University of Science and Technology of China（中国科学技术大学）
