R-LAM: Reproducibility-Constrained Large Action Models for Scientific Workflow Automation

TL;DR

R-LAM introduces a reproducibility-focused framework for scientific workflow automation using Large Action Models, ensuring deterministic, auditable, and replayable actions to enhance reliability and reproducibility in scientific research.

Contribution

The paper presents R-LAM, a novel framework that constrains Large Action Models with structured schemas and policies to meet scientific reproducibility requirements.

Findings

Improves reproducibility success rates in scientific workflows.

Enhances execution reliability over unconstrained LAM agents.

Supports iterative experimentation with controlled workflow forking.

Abstract

Large Action Models (LAMs) extend large language models by enabling autonomous decision-making and tool execution, making them promising for automating scientific workflows. However, scientific workflows impose strict requirements on reproducibility, auditability, and deterministic execution, which are not satisfied by generic LLM-based agents. Unconstrained action generation can lead to silent state changes, non-deterministic executions, and irreproducible experimental results, limiting the applicability of LAMs in scientific settings. In this paper, we propose R-LAM, a reproducibility-constrained framework for applying Large Action Models to scientific workflow automation. R-LAM introduces structured action schemas, deterministic execution policies, and explicit provenance tracking to ensure that every action and intermediate artifact is auditable and replayable. The framework supports failure-aware execution loops and controlled workflow forking, enabling iterative experimentation without compromising reproducibility. We implement R-LAM as a lightweight Python framework and release it as an open-source PyPI package to facilitate reproducible research. An experimental evaluation of representative scientific workflows demonstrates that R-LAM improves reproducibility success rates and execution reliability compared to unconstrained LLM-based agents, while retaining adaptive control over workflow execution.