Balancing Autonomy and Oversight in Language-Agent-Guided Physical Experimentation

<title>Abstract</title> <p> The rapid discovery of new materials for energy storage, advanced computing, and other technologies requires us to remake scientific experimentation beyond the limits of human cognition and manual interaction. Most existing autonomous experimentation relies on deterministic scripts or narrowly-scoped post-hoc machine learning. Large language models offer a fundamentally different capability, in the form of generalized reasoning from natural language goals, but we must still address safe LLM reasoning with real-time instrument feedback in a closed-loop workflow. Here we present <bold>A</bold> utonomous <bold>D</bold> ecision-Making <bold>A</bold> gent for <bold>M</bold> aterials Experimentation (ADAM), a multi-agent large language model framework integrating a retrieval augmented generation pipeline grounded in internal documentation, a validated instrument tool registry, and human-in-the-loop oversight positioned at semantic reversibility boundaries. By inferring these boundaries from natural language rather than hard-coded rules, ADAM achieves high autonomy in routine experiments and applies precise human oversight only when needed. ADAM provides a scalable blueprint for safe, deployment-ready autonomous experimentation and demonstrates how to reconcile autonomy and oversight in frontier models for scientific reasoning and experimentation. </p>