Junyao Yang

Junyao Yang

he/his

Junyao Yang's Homepage

RewardDS: Privacy-Preserving Fine-Tuning for Large Language Models via Reward Driven Data Synthesis

Jianwei Wang, Chengming Shi, Junyao Yang, Haoran Li, Qianli Ma, Huiping Zhuang, Cen Chen, Ziqian Zeng†

† Corresponding author

RewardDS synthetic-data generation, reward-guided filtering, refinement and fine-tuning pipeline

Reward proxy models guide the filtering and self-optimizing refinement of synthetic data for privacy-preserving LLM fine-tuning. This reduces noise in generated supervision while protecting sensitive training data.

Paper arXiv Code