I hope you’re not actually trying to compare human cognitive processes to an LLM, because that would be incredibly reductive, not to mention lacking in any empirical grounding.