3 ms·
That doesn’t mean it does any “reasoning”. It generates a response text that looks like a response to the input text. The whole point is that it generates respo
by kahnclusions 3y ago
That doesn’t mean it does any “reasoning”. It generates a response text that looks like a response to the input text. The whole point is that it generates responses that aren’t in the training set. But the fact that the response is actually a correct one is just coincidence.
- a_wild_dandan 3y agoSo when a model succeeds on a reasoning task it's not positive evidence, but when it fails it's negative evidence? The good ol' confirmation bias feedback loop!