Observability & Evals
Introducing LangSmith Tuned Evaluators, starting with Perceived Error
J. Broekhuizen, S. KarkhanisAugust 18, 20265 min
Introducing LangSmith Tuned Evaluators, starting with Perceived Error
Introduction
How do you know if an agent did a "good" job? Standard accuracy metrics fail when evaluating open-ended, complex tasks. Enter Tuned Evaluators.
