4 ms·
I wouldn’t even say it’s failing. Recognizing and correcting obvious errors in speech is a feature not a bug, both in humans and LLMs. Is it more likely that th
by digbybk 3y ago
I wouldn’t even say it’s failing. Recognizing and correcting obvious errors in speech is a feature not a bug, both in humans and LLMs. Is it more likely that they meant to say feather twice or they mistyped?
- transcriptase 3y agoTrue, but would it not make more sense to ask rather than assume and provide a long winded explanation? And even if it was a typo and I meant to say rocks, the model will often assume a variety of metals and answer with those.
- qup 3y agoHow it proceeds in the case of human error is going to be almost totally removed from its "intelligence," and almost totally related to its training.
- digbybk 3y agoI’m not claiming otherwise here. I’m claiming that it’s not a good test because a human would (almost certainly correctly) also skip past the double feather and answer the likely spirit of the question.
- throwuwu 3y agoA human would probably say “did you mean a kg of feathers and a kg of lead?”