And what do you think Jev confidence score is?
replyHere's a hint: confidence is not generated by a model.
Maybe I'm missing something, but why couldn't it be generated by the model? In older classification tasks with transformers like BERT, you could absolutely obtain a confidence score.
replyJev API returns both confidence and probabilities.
replyBut confidence value is just a function applied to probabilities. It is not coming from the model, and it carries no additional information.
It is documented btw, and yet you will see plenty of claims that Jev is better than LLM because it returns both.
Thanks, fixed my understanding!
replyDo you think though that Luna being a model post-trained for chat produces over-confidence in logprobs?
Yeah, but I wouldn't be surprised OpenAI's decision API is a post-trained Luna with confidence calibration.
replyTypesafe claims that Jev is calibrated, but there are plenty of examples where it completely fails (predicting die roll being the most obvious one).
Unfortunately calibration is hard to benchmark.
The dice roll prediction is about the way the prompt is setup misunderstanding how Jev works (they treat the confidence score as a probability score, which it isn't).
replyIf you instead give it a list of probability for each number and ask it whats the probability of each number, the result will be accurate.
What’s the difference?
reply