Study finds rapid LLM development is outpacing research meant to explain the models

The analysis warns that scientific understanding of large language models is falling behind their deployment.