Comment by mmoll

It doesn’t mean that these “thoughts” influenced their final decision the way they would in humans. An LLM will tell you a lot of things it “considered” and its final output might still be completely independent of that.