When LLMs Invent Rust Crates: An Empirical Study of Hallucination Patterns and Mitigation

Jieming Zheng, Hao Guan, Yepang Liu

Large Language Models (LLMs) have become powerful tools for code generation, yet they remain prone to hallucinations$\unicode{x2013}$producing plausible but incorrect or fabricated…