AI Makes Human-Like Reasoning Mistakes

Large language models (LMs) can complete abstract reasoning tasks, but they are susceptible to many of the same types of mistakes made by humans. Andrew Lampinen, Ishita...