Short thread on the latest paper led by @joseluiz.bsky.social w/ @lawstuff.bsky.social: arxiv.org/abs/2503.00992 The paper addresses two issues w/ previous machine psychology papers (including our own): 1) are LLMs mastering concepts, or are they memorizing the data? 1/14
Evidence of conceptual mastery in the application of rules by Large Language Models
In this paper we leverage psychological methods to investigate LLMs' conceptual mastery in applying rules. We introduce a novel procedure to match the diversity of thought generated by LLMs to that ob...
arxiv.org