UIAAAV1: UiPath Agentic Automation Associate Exam

By UiPath · Question Mode

QuestionQ53

VerifiedSolution Design

You are designing an evaluation for an agent. The test should compare the agent’s free-text summary of a ticket against an expected human summary, rewarding responses that are semantically equivalent even when the wording differs. Which evaluator type best fits this criterion?

Explanation

A semantic-similarity evaluator compares the meaning of free-text output with a reference, so equivalent ticket summaries can receive credit despite different wording. Exact string comparison requires matching text, while JSON similarity is intended for structured JSON data. OpenAI documents text-similarity graders as grading text against a reference using similarity metrics.

Learn more

Community Discussion

No comments yet. Be the first to start the discussion!