This probably means that the answers from LLMs about the reasoning are close to how humans would answer if they were not really paying attention to what they said (and thus formed no short-term memory of it), as explained in Humans Who Are Not Concentrating Are Not General Intelligences.
This probably means that the answers from LLMs about the reasoning are close to how humans would answer if they were not really paying attention to what they said (and thus formed no short-term memory of it), as explained in Humans Who Are Not Concentrating Are Not General Intelligences.