El Agente Potente: High-Throughput Agentic Atomistic Simulations
Abstract
Foundational machine-learning interatomic potentials (MLIPs) are transforming atomistic simulations by achieving near-ab initio accuracy across large chemical spaces at a fraction of the computational cost. A central challenge in using these tools for high-throughput property calculations is translating high-level scientific intent into adaptive simulation campaigns without compromising workflow rigour. We introduce El Agente Potente, an agentic system that combines typed execution graphs with a complementary coding mode for MLIPs-driven atomistic simulations. Typed execution graphs provide structured and provenance-aware execution for standardized workflows, with large language models (LLMs) restricted to planning and routing while deterministic Python components perform scientific computation and validation. Complementing this structured execution, a coding agent constructs customized workflows for tasks requiring greater procedural flexibility while invoking existing Potente functions for supported calculations. We demonstrate El Agente Potente across computational materials discovery, molecular energy-landscape exploration, adsorption, and catalytic reaction workflows, together with systematic benchmarks of reproducibility and LLM token cost. These results establish typed execution graphs and code-based workflow construction as complementary mechanisms for agentic scientific computing, combining controlled, auditable execution with the flexibility required for customized atomistic simulations
Related
- El Agente Quntur: Agentic Active-Space Selection for CASSCF
- La Agente Óptima: Towards Agentic Self-Driving Laboratories
- PLUMED-JAX: GPU-native enhanced sampling for atomistic simulations
- Revealing the Deactivation Mechanism of Aconitase at the Atomistic Level by Multiscale Simulations
- Agentic Workflow for High-Throughput Synthesis Campaigns: Mapping the Persistent Phosphor Ca2ZnSi2O7:Eu2+,Nd3+
BibTeX
@misc{ko2026agente,
title = {{El Agente Potente: High-Throughput Agentic Atomistic Simulations}},
author = {Tsz Wai Ko and Jiaru Bai and Thomas Swanick and Yeonghun Kang and Changhyeok Choi and Angelina Qihong Jiang and Aiwei Yin and Varinia Bernales and Alán Aspuru-Guzik},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.14840},
eprint = {2609.14840},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.14840},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - El Agente Potente: High-Throughput Agentic Atomistic Simulations
AU - Tsz Wai Ko
AU - Jiaru Bai
AU - Thomas Swanick
AU - Yeonghun Kang
AU - Changhyeok Choi
AU - Angelina Qihong Jiang
AU - Aiwei Yin
AU - Varinia Bernales
AU - Alán Aspuru-Guzik
PY - 2026
DA - 2026/09/13
PB - arXiv
DO - 10.48550/arxiv.2609.14840
UR - https://arxiv.org/abs/2609.14840
AB - Foundational machine-learning interatomic potentials (MLIPs) are transforming atomistic simulations by achieving near-ab initio accuracy across large chemical spaces at a fraction of the computational cost. A central challenge in using these tools for high-throughput property calculations is translating high-level scientific intent into adaptive simulation campaigns without compromising workflow rigour. We introduce El Agente Potente, an agentic system that combines typed execution graphs with a complementary coding mode for MLIPs-driven atomistic simulations. Typed execution graphs provide structured and provenance-aware execution for standardized workflows, with large language models (LLMs) restricted to planning and routing while deterministic Python components perform scientific computation and validation. Complementing this structured execution, a coding agent constructs customized workflows for tasks requiring greater procedural flexibility while invoking existing Potente functions for supported calculations. We demonstrate El Agente Potente across computational materials discovery, molecular energy-landscape exploration, adsorption, and catalytic reaction workflows, together with systematic benchmarks of reproducibility and LLM token cost. These results establish typed execution graphs and code-based workflow construction as complementary mechanisms for agentic scientific computing, combining controlled, auditable execution with the flexibility required for customized atomistic simulations
ER -