Table 2.

Summary of the free parameters used in each model to predict rats' choices

	No. of parameters	α ₁	α ₂	κ ₁	κ ₂	σ _α	σ _κ
Standard Q (const)	2	0.45	0.00^a	1.25	0.00^a
F-Q (const)	3	0.55		1.25	0.55
DF-Q (const)	4	0.60	0.45	1.20	0.45
Local matching law	1	0.01		1.00^a	0^a
Standard Q (variable)	2					0.005	0.10
F-Q (variable)	2					0.005	0.07
DF-Q (variable)	2					0.010	0.07

The parameters were determined so that the likelihood of the training data was maximized. const, Constant.
↵^aIndicates fixed parameters.