3 ms·
>Given any particular example, LLMs might hallucinate more and a human might do a better job at accuracy This drastically depends on the example. For average t
by sigmoid10 1y ago
>Given any particular example, LLMs might hallucinate more and a human might do a better job at accuracy
This drastically depends on the example. For average trivia questions, modern LLMs (even smaller, open ones) beat humans easily.