Yunke Wang

With the growing potential of large language models (LLMs) in the legal domain, domain-specific finetuning and retrieval-augmented generation (RAG) methods have received widespread attention. However, current methods still suffer from hallucination risk and failing to resolve semantic drift and a...