Similar Items: The impact of critique on LLM-based model generation from natural language: the case of activity diagrams