4 ms·

I mean they could just feed the solutions into the training data. Then suddenly the bot will do real good at HLE.

• mbesto 6mo ago
  Exactly. This is called overfitting and it's most definitely a thing.