Comment by jonnycoder

How does this intersect with evaluation in LLM integration & testing?