Jenn J.

Vector similarity search threshold, synthetic data generation, LLM-as-a-judge, and frameworks Jenn J. · Follow Published in Towards Data Science · 10 min read · 1 hour ago — Evaluating a Retrieval-Augmented Generation (RAG) model is much easier when you have ground truth data against which to compare. But what if you don’t? That’s where things get a bit trickier. However, even in the absence of ground truth, there are still ways to assess how well