Saved in:
Bibliographic Details
Main Authors: Dawson, William, Beal, Louis, Curé, Yoann, Fisicaro, Giuseppe, Rolland, Dorian, Genovese, Luigi
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2604.22571
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914504891170816
author Dawson, William
Beal, Louis
Curé, Yoann
Fisicaro, Giuseppe
Rolland, Dorian
Genovese, Luigi
author_facet Dawson, William
Beal, Louis
Curé, Yoann
Fisicaro, Giuseppe
Rolland, Dorian
Genovese, Luigi
contents Large language models (LLMs) and agentic systems have recently demonstrated potential for automating scientific workflows, including atomistic simulations. However, their deployment in high-performance computing (HPC) environments remains limited by the lack of mechanisms ensuring correctness, reproducibility, and safe interaction with computational resources. Generated workflows suffer from inconsistencies, incorrect API usage, or invalid physical configurations - leading to failed or unreliable simulations. In this work, we introduce LARA-HPC, a validation-driven agentic framework to enable reliable workflow generation for atomistic modeling on HPC systems. Our approach is based on three key components: (i) a controlled execution layer that mediates all interactions with HPC resources; (ii) simulation-native validation through dry-run capabilities, enabling execution-level verification without incurring resource cost; and (iii) a multi-phase agentic pipeline combining retrieval-augmented generation and iterative refinement. We demonstrate the effectiveness of this approach performing an end-to-end atomistic simulation workflow on HPC by applying LARA-HPC to Density Functional Theory simulations. The results show that validation-driven generation significantly improves robustness and enables iterative correction of both syntactic and physical inconsistencies. More broadly, this work advocates for a shift from generation-first to validation-first paradigms in Artificial Intelligence (AI) assisted scientific computing. We argue that the future task of the computational physics community is to develop domain specific agentic systems based on structured tooling to realize an HPC enabled co-piloted research ecosystem.
format Preprint
id arxiv_https___arxiv_org_abs_2604_22571
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle LARA: Validation-Driven Agentic Supercomputer Workflows for Atomistic Modeling
Dawson, William
Beal, Louis
Curé, Yoann
Fisicaro, Giuseppe
Rolland, Dorian
Genovese, Luigi
Computational Physics
Large language models (LLMs) and agentic systems have recently demonstrated potential for automating scientific workflows, including atomistic simulations. However, their deployment in high-performance computing (HPC) environments remains limited by the lack of mechanisms ensuring correctness, reproducibility, and safe interaction with computational resources. Generated workflows suffer from inconsistencies, incorrect API usage, or invalid physical configurations - leading to failed or unreliable simulations. In this work, we introduce LARA-HPC, a validation-driven agentic framework to enable reliable workflow generation for atomistic modeling on HPC systems. Our approach is based on three key components: (i) a controlled execution layer that mediates all interactions with HPC resources; (ii) simulation-native validation through dry-run capabilities, enabling execution-level verification without incurring resource cost; and (iii) a multi-phase agentic pipeline combining retrieval-augmented generation and iterative refinement. We demonstrate the effectiveness of this approach performing an end-to-end atomistic simulation workflow on HPC by applying LARA-HPC to Density Functional Theory simulations. The results show that validation-driven generation significantly improves robustness and enables iterative correction of both syntactic and physical inconsistencies. More broadly, this work advocates for a shift from generation-first to validation-first paradigms in Artificial Intelligence (AI) assisted scientific computing. We argue that the future task of the computational physics community is to develop domain specific agentic systems based on structured tooling to realize an HPC enabled co-piloted research ecosystem.
title LARA: Validation-Driven Agentic Supercomputer Workflows for Atomistic Modeling
topic Computational Physics
url https://arxiv.org/abs/2604.22571