2 ms·

There's not really any reason to think this other than shibboleths, it's become a thought-deadening cliche. I've played with true base models at big tech over

There's not really any reason to think this other than shibboleths, it's become a thought-deadening cliche. I've played with true base models at big tech over a period of 6+ months, it's a much much more frustrating experience, not an "unlocked" or "honest" version. Setting temperature to 1.2 is a reasonable miniature taste of the vibe of it. It would be interesting to see it done using logit probabilities, I.e. instead of asking for p 0 to 1, instead, ask it true or false, set logit bias such that those are the only two tokens that could be output, and use the logit probability emitted by the API.

gwern 3y ago
> There's not really any reason to think this other than shibboleths, it's become a thought-deadening cliche. The official OpenAI GPT-4 paper specifically reporting that the RLHF model is badly uncalibrated, while the GPT-4 base mode is pretty well calibrated, is a good reason to think that, and is not a 'thought-deadening cliche shibboleth': https://arxiv.org/pdf/2303.08774.pdf#page=12 https://arxiv.org/pdf/2303.08774.pdf#page=12 "The post-training hurts calibration significantly."
- refulgentis 3y ago
  Fascinating - thank you!!