Skip to main content
In this example, the generator creates output and passes it to the evaluator, which evaluates the response. If the evaluation fails, the evaluator returns corrections, and the generator is called again using the corrected output.
In response to the prompt, our agents generate this response: