LLM language comprehension task tests reveal insensitivity to underlying meaning 1 points by philipswood 4 hours ago 1 comments story
Hacker News