Thomas H. B. FitzGerald

The paper proposes finite retrospective inference as a bounded rationality approach to improve inference and learning by limiting the number of past states considered.

The paper presents two computational mechanisms for goal-directed exploration based on different types of information-gain, showing how they induce distinct patterns of Bayes-optimal behavior.

This study uses transcranial magnetic stimulation to show that angular gyrus selectively impairs free recall of autobiographical memories but not cued recall or word pair memories.