Table 2.

Group-level results of Bayesian K-fold cross validation.

Computational model	Parameter	elpd		Δelpd
wP-RL	8	− 37,412	(782)
P-RL	7	− 37,573	(785)	− 161	(23)
MB-RL	4	− 37,713	(797)	− 301	(46)
AU	4	− 40,269	(805)	− 2,857	(118)

Parameter number of free parameters, elpd estimated log predictive density, Δelpd difference in estimated log pointwise predictive density between a model and the best performing model; standard error in parentheses, AU attentional-updating model, MB-RL only model-based reinforcement-learning model, P-RL parallel reinforcement-learning model, wP-RL weighted parallel reinforcement-learning model.