The case for zero-error horizons in trustworthy LLMs (arxiv.org)

by daigoba66 • 116 comments • 79 points