Skip to main content
. 2014 Jul 8;470(2167):20130828. doi: 10.1098/rspa.2013.0828

Figure 8.

Figure 8.

Results of experiment 3 when an inferred reward function that includes information gain is used to estimate the prior distribution over Boolean concepts (r(18)=0.82, p<0.001). Each circle represents a game, and the least-squares regression line is shown.