This is excellent. It reminds me of theoretical vs experimental physics. Actual experiments to probe what is going on in the black box seem unintuitive to me and I really appreciate when someone can explain it so clearly. Interpretability is going to reveal so much about our minds and the machine minds.
This is excellent. It reminds me of theoretical vs experimental physics. Actual experiments to probe what is going on in the black box seem unintuitive to me and I really appreciate when someone can explain it so clearly. Interpretability is going to reveal so much about our minds and the machine minds.