Conceptoutput consistency check
Output Consistency Check
A method for deciding whether to request help that requires no additional trained classifier: the same small model is sampled multiple times, and confidence is judged from the consistency of its answers, which determines whether a larger model is consulted.
Timeline (last 90 days)
Nov 6
Output consistency checking is a method for deciding whether to ask for help that requires no additional classifier training: have the same small model sample answers to the same question multiple times; if the results are highly consistent, the model is considered confident about the question; if the results diverge, escalation is triggered.
Unverified50%