用多模态大模型实时管理智能体记忆，提升复杂任务完成率。
MemCtrl: Using MLLMs as Active Memory Controllers on Embodied Agents
- 让多模态大模型自带记忆筛选器，动态决定保留或丢弃信息。
- 在多个任务集上平均提升16%完成率，特定指令提升超20%。
- 适合资源受限的在线智能体，尤其擅长处理长而复杂的指令。
基础模型依赖上下文学习实现个性化决策，但上下文窗口有限，需借助RAG等记忆压缩与检索系统。现有系统通常将记忆视为离线存储，不利于受严格内存与计算约束的在线运行的具身智能体。本文提出MemCtrl框架，利用多模态大语言模型（MLLM）在线进行记忆剪枝。MemCtrl通过可训练的记忆头μ作为门控机制，决定探索过程中哪些观测或反思应保留、更新或丢弃。我们通过离线专家和在线强化学习两种方式训练μ，结果显示μ增强的MLLM在具身任务完成能力上显著提升。在EmbodiedBench多个子集上，对两个性能较差的MLLM进行μ增强后，平均提升约16%，部分指令子集提升超过20%。此外，对μ收集的记忆片段进行定性分析，发现其在长且复杂指令任务中表现更优。
原文摘要 · Abstract (English)
Foundation models rely on in-context learning for personalized decision making. The limited size of this context window necessitates memory compression and retrieval systems like RAG. These systems however often treat memory as large offline storage spaces, which is unfavorable for embodied agents that are expected to operate under strict memory and compute constraints, online. In this work, we propose MemCtrl, a novel framework that uses Multimodal Large Language Models (MLLMs) for pruning memory online. MemCtrl augments MLLMs with a trainable memory head μthat acts as a gate to determine which observations or reflections to retain, update, or discard during exploration. We evaluate with training two types of μ, 1) via an offline expert, and 2) via online RL, and observe significant improvement in overall embodied task completion ability on μ-augmented MLLMs. In particular, on augmenting two low performing MLLMs with MemCtrl on multiple subsets of the EmbodiedBench benchmark, we observe that μ-augmented MLLMs show an improvement of around 16% on average, with over 20% on specific instruction subsets. Finally, we present a qualitative analysis on the memory fragments collected by μ, noting the superior performance of μaugmented MLLMs on long and complex instruction types.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。