Sensitivity of Small Language Models to Fine-tuning Data Contamination

Open in new window