Appendix 2—figure 1. Comparison of actual and recovered parameters via ‘BAYES_π’ model.

θ_p and θ_n - positive and negative modes of the bimodal distribution of prior expectations, σ_exp - uncertainty of the prior distribution, σ_sens uncertainty in the sensory likelihood, α - fraction of random estimations, π - mixing parameter responsible for the degree of bimodality. Actual parameters are scattered along x-axis and recovered parameters are scattered along y-axis. The dashed diagonal line is a reference line indicating perfect parameter recovery. Pearson’s correlation coefficients are indicated above each plot.