Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

Does that mean human brains have neuron "training" feedback?


There's a few theories.

Most fundamentally come down to the neurones connections and behaviour can be modelled as minimizing prediction error in their stimulus.

Dopamine encoding an outcome compared to a baseline expectation is very close to tempeoal difference learning.

This is basically what slot machines exploit with their variable reward schemes.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: