作者位次 #3/6ACL 2026Poster· Natural Language Processing
S tealth G raph: Exposing Domain-Specific Risks in LLM s through Knowledge-Graph-Guided Harmful Prompt Generation
Huawei Zheng · Xinqi Jiang · Sen Yang · Shouling Ji · Yingcai Wu · Dazhen Deng
摘要
Large language models (LLMs) are increasingly applied in specialized domains such as finance and healthcare, where they introduce unique safety risks. Domain-specific datasets of harmful prompts remain scarce and still largely rely on manual construction; public datasets mainly focus on explicit...