3 ms·
Not true or false; just present or absent in the reference data. Note that false negatives will not result in erroneous output, so the model can safely err on t
by mortehu 4y ago
Not true or false; just present or absent in the reference data. Note that false negatives will not result in erroneous output, so the model can safely err on the side of caution.
Also 100% accuracy is probably not the real threshold for being useful. There are many low hanging fruits today that could be solved by absolutely tiny error correcting models (e.g. arithmetic and rhyming).
- astrange 4y agoThere's research showing you can tell if something is a hallucination or memorized fact based on the activation patterns inside the LM.