3 ms·
The mistake you make here is to forget that the training data of the original models was also _full_ or errors and biases — and yet they still produced coherent
by bigfudge 2y ago
The mistake you make here is to forget that the training data of the original models was also _full_ or errors and biases — and yet they still produced coherent and useful output. LLM training seems to be incredibly resilient to noise in the training set.