Problem
Research questions and friction points this paper is trying to address.
High-Performance Computing
Large Language Model
Validation-Gated Execution
Scientific Applications
Edge-Controlled
ECAS: An Edge-Controlled Agentic System for Validation-Gated Scientific Application Execution
🤖 AI Summary
为解决科学应用在高性能计算中自动化执行问题，提出ECAS系统，通过云模型推理、边缘控制和验证门控执行的方法，实现低人工干预下的闭环执行。
📝 Abstract
Scientific applications increasingly rely on high-performance computing (HPC), yet translating a scientist's high-level goal into a correct target-scale execution remains brittle and labor-intensive. Large language model (LLM) agents promise to automate this, but two obstacles remain: granting a cloud-hosted model direct HPC access exposes credentials and execution authority, while withholding it demands continuous human supervision; and one-shot generation cannot adapt when generated artifacts fail in a site-specific HPC environment. We present \textsc{ECAS}, an \textbf{E}dge-\textbf{C}ontrolled \textbf{A}gentic \textbf{S}ystem for closed-loop execution of scientific computing campaigns with limited human intervention. \textsc{ECAS} separates \emph{reasoning}, \emph{control}, and \emph{execution}: a cloud-hosted LLM proposes plans, artifacts, and repairs; a user-controlled edge agent retains credentials, workflow state, and execution authority while enforcing policy and resource constraints; and the HPC system computes. Its core mechanism is \emph{validation-gated execution}: generated artifacts pass static checks and small-scale validation, failures trigger repairs from sanitized execution feedback, and target-scale execution is permitted only after validation and policy checks pass. \textsc{ECAS} also draws on an edge-resident library of expert-distilled, site-specific skills that is never disclosed to the cloud. In preliminary experiments with three scientific applications on two production ALCF systems under six injected fault types, closed-loop repair improves application success from 0/6 to 6/6 over one-shot generation, validation gating prevents all three observed target-scale failures, and skill conditioning improves success from 4/6 to 6/6. These results show the feasibility of delegating adaptive reasoning to the cloud while retaining execution control at the edge.
Innovation
Methods, ideas, or system contributions that make the work stand out.
Edge-Controlled Agentic System
validation-gated execution
closed-loop repair
🔎 Similar Papers
No similar papers found.
