Large language models struggle with human-like reasoning shortcuts

Original: Tracing the ongoing emergence of human-like reasoning in Large Language Models

Source: arxiv.org ↗