Comment by tedsanders
17 hours ago
I intended no dismissiveness or condescension. My hope was to explain why it's hard to prove whether something affects model behavior. In the case of the moon, we have a strong prior belief that it makes no real difference. But it's hard to prove, because what if there's an unexpected impact from tides, cosmic rays, grid voltages, holiday traffic, etc. Models trained under slightly different conditions could have slightly different weights and behave slightly differently when solving math problems. Similarly, I have a strong expectation that, for example, a thumbs up signal from a ChatGPT chat will not meaningfully affect long-horizon mathematics work in our latest model, but it's always possible that it could. I think the plausibility of the ChatGPT route is higher than the tides, but still incredibly low. I respect Tristan and Levant a great deal and I'm bummed that this controversy has erupted (I acknowledge this will ring hollow if you think it's our fault). It reminds me a bit of the Frontier Math controversy, where people on the internet boldly claimed over and over again that we had trained on the Frontier Math evaluation set, even though we had not.
We aren’t dummies, we know it’s hard to prove exactly how significant of an impact that would have on the result. Nobody expect you to do that. There are a lot of steps and things that are possible to check _before_ the need for such a strict definition of „proof“
You seem to jump over the principal issue of whether any data from the researchers used to train or otherwise affect the model which produced the OpenAI proof.
We can judge for ourselves the impact and degree of that wrongdoing, but it seems OpenAI is confirming: yes, that is what happened, but with more words.