Different betas for acting vs observing

maxkw / evolution

0 stars 0 forks source link

Closed polyguo closed 7 years ago

polyguo commented 7 years ago

our agents should expect others to be less noisy than they are this translates to the beta used for ToM being higher than the one used when acting

maxkw commented 7 years ago

didn't implement. beta is an exploration parameter so it should be treated as part of theory of mind.