Agentic AI -- Physicist Collaboration in Experimental Particle Physics: A Proof-of-Concept Measurement with LEP Open Data

TL;DR

This paper demonstrates AI agents performing an experimental physics measurement on LEP data, showcasing a step toward automated theory-experiment collaboration in fundamental physics.

Contribution

It introduces a proof-of-concept where AI agents autonomously conduct a particle physics measurement, integrating data analysis and note writing under physicist supervision.

Findings

AI agents successfully measured the thrust distribution in electron-positron collisions.

The analysis was fully carried out by AI agents, including data correction and note writing.

This work illustrates the potential for AI to accelerate scientific discovery in physics.

Abstract

We present an AI agentic measurement of the thrust distribution in $e^{+}e^{-}$ collisions at $\sqrt{s}=91.2$~GeV using archived ALEPH data. The analysis and all note writing is carried out entirely by AI agents (OpenAI Codex and Anthropic Claude) under expert physicist direction. A fully corrected spectrum is obtained via Iterative Bayesian Unfolding and Monte Carlo based corrections. This work represents a step toward a theory-experiment loop in which AI agents assist with experimental measurements and theoretical calculations, and synthesize insights by comparing the results, thereby accelerating the cycle that drives discovery in fundamental physics. Our work suggests that precision physics, leveraging the open LEP data and advanced theoretical landscape, provides an ideal testing ground for developing advanced AI systems for scientific applications.