Kierston: City-Scale Simulation for Criminal Investigation Benchmarking
November 2, 2026
Kierston is a city-scale simulation of Allentown, PA, generating telecommunications, financial, and administrative records for 125,000 residents. KierstonBench uses these simulated events to evaluate LLM capabilities in surveillance and investigative tasks.
HOW THIS AFFECTS YOU
●
researcherYou can use this benchmark to test long-horizon investigative reasoning with access to ground-truth simulated data.
●
policyThis provides a controlled environment to study the safety and accuracy of LLMs in law enforcement contexts.