作者位次 #6/7ACL 2026Poster· Natural Language Processing
Fingerprinting LLM s via Prompt Injection
Yuepeng Hu · Zhengyuan Jiang · Mengyuan Li · Osama Ahmed · Zhicong Huang · Cheng Hong · Neil Zhenqiang Gong
摘要
Large language models (LLMs) are often modified after release through post-processing such as post-training or quantization, which makes it challenging to determine whether one model is derived from another. Existing provenance detection methods have two main limitations: (1) they embed signals i...