LEMCA：基于大语言模型引导的高效模式切换控制架构合成
LEMCA: LLM-Guided Synthesis of Efficient Mode-Switching Control Architectures
浏览论文内容
中文总结 AI 辅助
针对传统模式切换控制器设计繁琐且单一架构资源浪费的问题，提出LEMCA方法，利用大语言模型在进化循环中迭代优化控制模式与资源分配，实验证明其能超越单一化设计的帕累托前沿，实现高效资源利用。
中文摘要 AI 辅助
自然世界中的物理控制任务，如驾驶或物体操作，其感官和计算复杂度随时间经常表现出剧烈变化。相应地，一种自然的资源高效型机器人控制选择是在具有不同资源分配的控制模式之间动态切换。然而，这类“模式切换控制器”（MSC）历来需要针对每个新任务进行繁琐且由专家驱动的设计与合成。受这些设计难题驱动，现代机器人控制架构常常退回到一种浪费的“单一化”一刀切结构，其中资源分配永久锚定于最困难、资源最密集的任务阶段。为了促进高性能且高效MSC的设计，我们提出了基于大语言模型引导的高效模式切换控制架构合成（LEMCA）。LEMCA将MSC设计表示为可解释的程序，并在进化循环中迭代细化。为了评估设计适应性，我们提出了与MSC兼容的自动控制器合成方法的扩展，例如模拟中的强化学习。随后，LEMCA利用大语言模型（LLM）的语义先验、推理和编码能力，迭代编辑控制模式、其对应的感官计算资源分配以及模式转换。我们在多个控制基准上的实验表明，LEMCA始终能发现超越单一化设计帕累托前沿的策略，通过在“简单”任务阶段回收浪费的资源。因此，LEMCA提供了一条自动化、低成本的路径来合成资源高效的MSC设计。
英文摘要
Physical control tasks in the natural world, such as driving or object manipulation, frequently exhibit dramatic variations in sensory and compute complexity over time. Correspondingly, a natural resource-efficient choice for robot control is to dynamically switch between control modes with varying resource allocations. However, such "mode-switching controllers" (MSCs) have historically required laborious, expert-driven design and synthesis for each new task. Driven by these design difficulties, modern robotic control architectures often fall back to a wasteful "monolithic" one-size-fits-all structure, where resource allocation is permanently anchored to the hardest, most resource-intensive task phases. To facilitate the design of performant yet efficient MSCs, we propose LLM-Guided synthesis of Efficient Mode-Switching Control Architectures (LEMCA). LEMCA represents MSC designs as interpretable programs to be iteratively refined in an evolutionary loop. To evaluate design fitness, we propose MSC-compatible extensions of automated controller synthesis approaches, such as reinforcement learning in simulation. LEMCA then leverages the semantic priors, reasoning, and coding capabilities of Large Language Models (LLMs) to iteratively edit controller modes, their corresponding sensory-compute resource allocations, and mode transitions. Our experiments across diverse control benchmarks show that LEMCA consistently discovers strategies that surpass the Pareto frontier of monolithic designs by reclaiming wasted resources during "easy" task phases. LEMCA thus presents an automated, low-effort path to synthesize resource-efficient MSC designs.
发表机构
- University of Pennsylvania（宾夕法尼亚大学）
- Amazon Robotics（亚马逊机器人）
- Georgia Institute of Technology（佐治亚理工学院）
机构由 AI 辅助整理，请以论文原文为准。
