Dataset for "Revealing how uncertainty computations drive hierarchical reasoning via CogLink Networks"

We run CogLink Network, its various perturbed version and other machine learning algorithms in various bandit tasks to understand the computational roles each neural mechanism serves in hierarchical reasoning under uncertainty. We show that these mechanistic features in CogLink Network enable efficient and flexible environmental exploration while replicating experimentally-observed neural signatures.

Furthermore, it shows utility in computational psychiatry, by linking prefrontal dysfunction to idiosyncratic reasoning in schizophrenia.