LLMs causing problems? The solution, apparently: more LLMs!
How would you know that the new paper contained the same content as the original (let alone the actual Lean code)?