arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2405.11458cs.AIcs.SYeess.SY
收藏

CPS-LLM:面向人在回路、人在现场的信息物理系统的基于大语言模型的安全使用计划生成器

CPS-LLM: Large Language Model based Safe Usage Plan Generator for Human-in-the-Loop Human-in-the-Plant Cyber-Physical System

Arizona State University(亚利桑那州立大学)

Ayan Banerjee, Aranyak Maity, Payal Kamboj, Sandeep K. S. Gupta

AI总结:

针对 LLM 在人在回路 CPS 中生成的计划可能不可行或不安全的问题,本文提出经指令微调的 CPS-LLM,融合物理动态系数估计与轨迹提示训练,可在自动胰岛素输注场景中生成安全可行计划。

摘要:

我们探索大语言模型(LLM)在人在回路、人在现场的信息物理系统(CPS)中的使用,以将高层提示转化为个性化行动计划,随后把该计划转换为对序贯决策的落地推理,并由真实世界的 CPS 控制器自动化执行,从而实现控制目标。我们表明,让 LLM 具备上下文以生成面向特定领域的计划相对直接。然而,这些计划对于物理系统而言可能不可执行,或者对人类用户可能不安全。为解决这一问题,我们提出 CPS-LLM,这是一种使用指令微调框架重新训练的 LLM,可确保生成的计划不仅符合 CPS 的物理系统动态,而且对人类用户安全。CPS-LLM 包含两个创新组件:a)基于液体时间常数神经网络的物理动态系数估计器,能够在存在部分不可测状态变量的情况下推导动态模型的系数;b)随后利用这些模型系数训练 LLM,其提示中嵌入了来自动态系统的轨迹及相应的模型系数。我们表明,当 CPS-LLM 与 BARD 等上下文化聊天机器人集成时,它能够生成可行且安全的计划,以管理外部事件,例如供 1 型糖尿病患者使用的自动胰岛素输注系统中的进餐事件。

英文摘要:

We explore the usage of large language models (LLM) in human-in-the-loop human-in-the-plant cyber-physical systems (CPS) to translate a high-level prompt into a personalized plan of actions, and subsequently convert that plan into a grounded inference of sequential decision-making automated by a real-world CPS controller to achieve a control goal. We show that it is relatively straightforward to contextualize an LLM so it can generate domain-specific plans. However, these plans may be infeasible for the physical system to execute or the plan may be unsafe for human users. To address this, we propose CPS-LLM, an LLM retrained using an instruction tuning framework, which ensures that generated plans not only align with the physical system dynamics of the CPS but are also safe for human users. The CPS-LLM consists of two innovative components: a) a liquid time constant neural network-based physical dynamics coefficient estimator that can derive coefficients of dynamical models with some unmeasured state variables; b) the model coefficients are then used to train an LLM with prompts embodied with traces from the dynamical system and the corresponding model coefficients. We show that when the CPS-LLM is integrated with a contextualized chatbot such as BARD it can generate feasible and safe plans to manage external events such as meals for automated insulin delivery systems used by Type 1 Diabetes subjects.

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。

↑