Liu, Ting

Towards Comprehensive Post Safety Alignment of Large Language Models via Safety Patching

2024 · arXiv (Cornell University)

Safety alignment of large language models (LLMs) has been gaining increasing attention. However, current safety-aligned LLMs suffer from the fragile and imbalanced safety mechanisms, which can still be induced to generate unsafe responses, exhibit over-safety …

FroM: Frobenius Norm-Based Data-Free Adaptive Model Merging

2025 · arXiv (Cornell University)

With the development of large language models, fine-tuning has emerged as an effective method to enhance performance in specific scenarios by injecting domain-specific knowledge. In this context, model merging techniques provide a solution for fusing …