Fig. 4: Effects of psilocybin on model parameters across days of reversal learning.
From: Psilocybin increases optimistic engagement over time: computational modelling of behaviour in rats
The first 5 panels in this figure illustrate pairwise comparisons of the model parameters across days including (A) forgetting rate for losses (logit-scaled), (B) forgetting rate for rewards (logit-scaled), (C) loss aversion (log-scaled), (D) action precision (log-scaled), and (E) prior reward probability (logit-scaled). Estimated marginal means are represented with dotted lines (psilocybin group) or solid lines (saline group) with shading indicating SEM. Solid red lines indicate significant differences between groups (p < 0.05 Correlations between parameters in the best-fit computational (AI) model and behavioural results measures (F). *p < 0.05, **p < 0.01, ***p < 0.001.
