PPL
Longtao Huang
Papers (4)
How LoRA Remembers? A Parametric Memory Law for LLM FinetuningMake LLM Learn to Synthesize from Streaming Experiences through FeedbackThe Bridge-Garden Dilemma in LLM Distillation: Why Mixing Hard and Soft Labels WorksLocalization then Neutralization: Gradient-guided Token Suppression against Visual Prompt Injection Attack