Large Model Application Algorithm Research Scientist-International Content Security Algorithm Research

Build and iterate on large-model approaches for international content safety, helping detect risks earlier, respond to incidents faster, and strengthen monitoring across content moderation, search, and recommendation. Focus on RL-based reasoning advancement, including designing reward models, stabilizing RL training without costly SFT data, extending reasoning methods to natural-language tasks, and improving reasoning efficiency via techniques such as knowledge distillation and Long-CoT.