Resolution Criteria

This market resolves NO if, by December 2029, an agentic AI system autonomously performs an academically rigorous analysis of LHC particle physics data that passes peer review by experimental particle physicists.

The Task

Given ~100 PB of LHC pre-reconstruction data and a prompt like "measure the Standard Model Higgs mass" or "evaluate this Beyond the Standard Model scenario," the agent must autonomously:

• Plan and execute the full analysis chain

• Perform low-level preprocessing/reconstruction on distributed computing

• Design analysis strategy (triggers, event selection, object definitions)

• Estimate backgrounds with full systematic uncertainties (experimental + theoretical)

• Conduct state-of-the-art statistical analysis

Constraints

• Agent has read access to internal documentation and low-level software

• No access to prior analysis code or detailed prior results for the same task

• Must use comparable compute resources to a human analysis team (excluding agent inference costs)

• Success = 1 out of 3 runs passes peer review