FermiLink: A Unified Agent Framework for Multidomain Autonomous Scientific Simulations

TL;DR

FermiLink is an open-source, unified agent framework that enables multidomain scientific simulations, reproducing figures and producing research-grade results across diverse scientific software packages.

Contribution

It introduces a modular framework separating package knowledge from workflows, supporting uniform operations across multiple scientific domains and demonstrating high-fidelity reproductions.

Findings

Reproduces 74 out of 132 figures with simulations

Achieves high-fidelity agreement in 30 cases

Produces research-grade results on unpublished problems

Abstract

Artificial-intelligence (AI) agent frameworks have been developed for autonomous scientific simulations, but most current agent frameworks are tailored to a single or a small set of software packages. Herein, FermiLink, a unified and extensible open-source agent framework is introduced for multidomain scientific simulations. Its key design principle is the separation of package knowledge bases from simulation workflows, so that simulation workflows in FermiLink, from figure-level simulations to full-paper-level research on high-performance computing clusters, operate uniformly among supported packages via a four-layer progressive disclosure mechanism. Using OpenAI Codex as the agent provider, the capabilities of FermiLink are demonstrated across approximately 50 scientific software packages spanning nine research domains from physics to engineering. Systematic benchmarks on 132 real-world figure-level reproduction tasks with 44 packages show that FermiLink reproduces 74 (56.1%) of published figures with simulations, among which 30 achieve high-fidelity agreement and 35 reach qualitative agreement with the target figures. A smaller set of human expert-guided reproduction benchmarks with 10 packages further highlights the importance of expert insights for improving the simulation fidelity. Beyond reproduction, a single-blinded study demonstrates that FermiLink can produce research-grade results on unpublished polariton physics problems when provided with sufficiently detailed research objectives and source code, even in the absence of external documentation or tutorials. Overall, FermiLink provides a scalable research infrastructure that may accelerate the path from scientific questions to computational results across diverse domains.