It really shows how LLMs work. It's all about probabilities, and not about understanding. If something looks very similar to a well known problem, the llm is ha

It really shows how LLMs work. It's all about probabilities, and not about understanding. If something looks very similar to a well known problem, the llm is having a hard time to "see" contradictions. Even if it's really easy to notice for humans.