Medical large language models are vulnerable to data-poisoning attacksnature.com2 points·gnabgib··0 commentsOpen articleSaveView on HN