LLM Safety From Within: Detecting Harmful Content with Internal Representations Paper • 2604.18519 • Published 28 days ago • 26
Adam's Law: Textual Frequency Law on Large Language Models Paper • 2604.02176 • Published Apr 2 • 503
