It doesnt seem possible to have a fixed positivity ratio when considering research on extinction. The gist of extinction is that when training using positive f

It doesnt seem possible to have a fixed positivity ratio when considering research on extinction. The gist of extinction is that when training using positive feedback, the desired response will cease when the positive feedback does (extinction). In order to maintain the behavior the feedback has to essentially be randomized with a decreasing prevalence (positivity ratio).