Hallucinated citations are polluting the scientific literature. What can be done?