The Rosetta Paradox: Domain-Specific Performance Inversions in Large Language Models

Basab Jha, Ujjwal Puri

While large language models, such as GPT and BERT, have already demonstrated unprecedented skills in everything from natural language processing to domain-specific applications, th…