Table 1.

Mean parameter fits ± SEM from 20 subjects for three models

	Models
	Q-learning		Policy-gradient		Logistic regression
QL,0	0.54 ± 0.09		—	Rc	3.44 ± 1.51
QR,0	0.61 ± 0.08	η	0.48 ± 0.06	Ru	2.09 ± 0.053
α	0.35 ± 0.07	α	62.23 ± 20.22	Repetition	1.89 ± 0.58
κ * α	0.24 ± 0.07	κ	0.48 ± 0.07	Intercept	−0.03 ± 0.06
β	9.08 ± 2.62	W0	−14.28 ± 35.87		—