From human explanations to explainable AI: Insights from constrained optimization

Inga Ibs, Claire Ott, Frank Jäkel, Constantin A. Rothkopf

Abstract: Highlights•We introduce a problem solving paradigm to study explanations for optimization.•We utilize an exploration and a sequential decision-making version of the task.•We derive formal strategies from explanations, verbal reports, and behavioral data.•The results provide insights for the generation of cognitively adequate explanations.