DuoMem: 通过双空间蒸馏实现高效的设备端记忆代理
DuoMem: Towards Capable On-Device Memory Agents via Dual-Space Distillation
AI总结：
提出DuoMem双空间蒸馏框架，将大型教师模型的过程问题解决能力转移到紧凑学生模型，在ALFWorld基准上将4B参数模型的任务成功率从4.3%提升至77.9%，接近72B教师模型的87.1%，且参数增加少于10M，速度提升3倍以上。
摘要：
基于大型语言模型（LLM）的代理可以通过与环境的多轮交互解决复杂的过程性任务，但这种能力通常依赖于大型模型、长上下文和重复推理调用。这使得先进的记忆增强代理难以部署在资源受限的设备上。我们引入了DuoMem，一个双空间蒸馏框架，将过程问题解决能力从大型教师模型转移到紧凑的学生模型。DuoMem在两个互补的空间中进行蒸馏：（1）上下文空间蒸馏，用更高质量的教师生成的过程记忆替换学生生成的记忆，并将其前置到学生输入中；（2）参数空间蒸馏，在成功的教师轨迹上微调轻量级LoRA适配器。在具有挑战性的具身决策基准ALFWorld上评估，DuoMem将4B参数模型的任务成功率从4.3%提升至77.9%，缩小了与72B教师模型（87.1%）的大部分差距，同时仅增加了不到10M可训练参数和几兆字节的预计算教师记忆。此外，DuoMem增强的4B模型在挂钟时间上完成任务的速度比72B教师模型快3倍以上，使其适用于实时边缘部署，而这对于教师模型来说将是困难的。跨8个模型（参数范围2B-72B）的广泛消融实验表明，两个蒸馏轴都贡献了互补的效果。
英文摘要：
Large Language Model (LLM)-based agents can solve complex procedural tasks by interacting with environments over multiple turns, but this ability typically depends on large models, long contexts, and repeated inference calls. This makes advanced memory-augmented agents difficult to deploy on resource-constrained devices. We introduce DuoMem, a dual-space distillation framework that transfers procedural problem-solving ability from a large teacher model to compact student models. DuoMem distils in two complementary spaces: (1)context-space distillation, which replaces student-generated memories with higher-quality teacher-generated procedural memories prepended to the student's input, and (2)parameter-space distillation, which fine-tunes lightweight LoRA adapters on successful teacher trajectories. Evaluated on ALFWorld, a challenging embodied decision-making benchmark, DuoMem boosts a 4B-parameter model from 4.3% to 77.9% task success rate, closing most of the gap to a 72B teacher model (87.1%), while adding fewer than 10M trainable parameters and only a few megabytes of pre-computed teacher memories. Moreover, the DuoMem-enhanced 4B model completes tasks over 3x faster than the 72B teacher in wall-clock time, making it viable for real-time edge deployment, which would be challenging for the teacher.Extensive ablations across eight models spanning 2B-72B parameters reveal that both distillation axes contribute complementary
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
