Coping with dumb LLMs using classic ML

from blog Doug Turnbull, | ↗ original
Taking the output of many dumb LLM search relevance judges and feeding the output to a decision tree to improve precision