Comment on Building reliable systems out of unreliable agentsparentComments−infecto2yFor clarification on the first part. The research suggests you can utilize the same prompt over multiple runs as the input to picking the answer.
Comments
For clarification on the first part. The research suggests you can utilize the same prompt over multiple runs as the input to picking the answer.