Xinfeng Li's picture

Xinfeng Li

LetterJohn

·

AI & ML interests

Trustworthy AI, AI for Security & Privacy

Recent Activity

upvoted a paper 2 days ago

Internal Safety Collapse in Frontier Large Language Models

upvoted a paper 23 days ago

BandPO: Bridging Trust Regions and Ratio Clipping via Probability-Aware Bounds for LLM Reinforcement Learning

upvoted a paper about 2 months ago

AgentDoG: A Diagnostic Guardrail Framework for AI Agent Safety and Security

View all activity

Organizations

LetterJohn 's models 1

LetterJohn/SafeGen-Pretrained-Weights

Text-to-Image • Updated Feb 17, 2025 • 18 • 2