Sensitivity to outcome devaluation in operant tasks is better predicted by food restriction level than reinforcement training schedule in mice https://www.biorxiv.org/content/10.1101/2023.02.23.529699v1?med=mas