On: DeepSeek-R1

[wat10000]

I’ll bet said PhDs will tell you they don’t know instead of confidently stating the wrong answer in this case. Getting LLMs to express an appropriate level of confidence in their output remains a major problem.