Fig. 6.

TTC estimation of exploration is sensitive to the balance between exploration and motor noise. We compare model behavior for an adaptive (a, b) and a random reward criterion (c, d). a, c Similarity ratio as a function of the ratio of exploration and motor noise variance parameters. The horizontal line at SR = 1 indicates perfect estimation of exploration. Symbols and other details as in Fig. 5. b, d Draws of motor noise and exploration on successful and non-successful trials in five simulations of a simulation set of the Therrien16 model, resulting in 2500 samples plotted in total for the rewarded and non-rewarded panel together. Except for the reward criterion, other parameters in the simulations have their default values. The data are split based on whether these trials were rewarded or not. Dotted lines indicate $\pm \sigma$