Table 1. Model comparison.

The “winning” model is the “Confirmation” model for which the learning rates are displayed in Fig 3C. The second best model is the Full model, for which the learning rates are displayed in Fig 2B.

Model	Full (5df)	Information (3df)	Valence (3df)	Confirmation (3df)	Perseveration (4df)	One (2df)
BIC	162.0±13.4	178.2±13.0	180.7±11.8	155.0±13.2	165.2±13.6	179.1±11.8
PP	0.02±0.02	0.00±0.00	0.05±0.05	0.89±0.06	0.01±0.01	0.04±0.03
XP	0.00	0.0	0.0	1.0	0.0	0.0

BIC: Bayesian Information Criterion; PP: posterior probability; XP: exceedance probability; df: degrees of freedom.