The Role of Fine-grained Harm Signals in LLM Safety
The findings show that even a direction orthogonal to a concept at one layer can contribute to the concept's downstream amplification, and demonstrate that more fine-grained category residuals should also be considered beyond shared general harmfulness representation to fully understand LLM safety.
Soyeon Park, Seogyeong Jeong, Sunwoo Kim et al.
· 0 citations