Table 1. Model specification.

Model	Parameters
Reward learning only	${\alpha }^r$		${\beta }^r$
Punishment learning only	${\alpha }^p$		${\beta }^p$
Symmetrical learning	$\alpha$		$\beta$
Asymmetric learning rates	${\alpha }^r$	${\alpha }^p$	$\beta$
Asymmetric sensitivities	$\alpha$		${\beta }^r$	${\beta }^p$
Asymmetric learning rates and sensitivities	${\alpha }^r$	${\alpha }^p$	${\beta }^r$	${\beta }^p$

All custom model scripts are available online at https://osf.io/m3zev/.