Paper

Agentic programs: an emerging form of scientific software in computational materials science

Computational materials science has traditionally delegated algorithmic tasks to computers while leaving scientific judgments to humans. We argue that recent LLM-based agent harnesses enable an emerging form of scientific software, agentic programs, that combine deterministic algorithms with bounded LLM-based judgment, task-specific verification, episodic maturation, and complete delegation in production. We illustra · Authors: Yunsung Lim, Haekwan Jeon, Jaesun Kim, Jisu Kim, Seungwu Han · Published 2026-09-01 · arXiv:2609.00795

Published
2026-09-01
Authors
5
arXiv
2609.00795
Category
cond-mat.mtrl-sci

Abstract

arXiv:2609.00795

Yunsung Lim, Haekwan Jeon, Jaesun Kim, Jisu Kim, Seungwu Han

Computational materials science has traditionally delegated algorithmic tasks to computers while leaving scientific judgments to humans. We argue that recent LLM-based agent harnesses enable an emerging form of scientific software, agentic programs, that combine deterministic algorithms with bounded LLM-based judgment, task-specific verification, episodic maturation, and complete delegation in production. We illustrate this concept with DeMARS, an agentic program for constructing atomistic models from experimentally measured disordered crystal structures.