PNASCounterfeit judgments in large language modelsMatjaž Perc2025-11-24Full textDOI https://doi.org/10.1073/pnas.2528527122