LSD increases the rate at which people learn from both rewards and punishments during a probabilistic reversal learning task, suggesting a state of heightened learning plasticity. Healthy volunteers given intravenous LSD or placebo completed a task where they had to learn which of three stimuli was most often rewarded, with the reward contingencies later reversing. Computational modeling of reinforcement learning showed that LSD primarily enhanced the reward learning rate and also elevated the punishment learning rate, while decreasing stimulus stickiness (a measure of choice repetition), indicating increased exploration. These effects point to a potential mechanism by which LSD could help revise maladaptive associations in clinical treatment.
Lysergic acid diethylamide (LSD) increases the speed at which the brain updates the value of actions following feedback, particularly after rewards, and makes behavior more exploratory. In a within-subjects experiment, healthy volunteers received intravenous LSD or placebo and performed a probabilistic reversal learning task where they learned which of three stimuli was most often rewarded, with contingencies later reversing. Computational modeling showed LSD enhanced the reward learning rate and also elevated the punishment learning rate, while reducing stimulus stickiness—a measure of choice repetition regardless of outcomes. Conventional measures of immediate feedback sensitivity were unaffected. These findings suggest LSD induces a state of heightened plasticity that may help revise maladaptive associations in clinical settings.