Li, C.; Flanigan, J. Task Contamination: Language Models May Not Be Few-Shot Anymore. AAAI 2024, 38, 18471-18480.