作者位次 #5/7ACL 2026Poster· Natural Language Processing
CTRAP : Embedding Collapse Trap to Safeguard Large Language Models from Harmful Fine-Tuning
Biao Yi · Tiansheng Huang · Baolei Zhang · Tong Li · Lihai Nie · Zheli Liu · Li Shen
摘要
Fine-tuning-as-a-service, while commercially successful for Large Language Model (LLM) providers, exposes models to harmful fine-tuning attacks. As a widely explored defense paradigm against such attacks, unlearning attempts to remove malicious knowledge from LLMs, thereby essentially preventing...