ITHACA, N.Y. – Large language models (LLMs) show promise as a tool for exploring the vast scientific literature, but are they trustworthy when it comes to providing full and scientifically accurate ...