Junxiao Wang

Private Training Large-scale Models with Efficient DP-SGD

Scalable Zeroth-Order Fine-Tuning for Extremely Large Language Models with Limited GPU Memory

FlashDP: Memory-Efficient and High-Throughput DP-SGD Training for Large Language Models

ZO-Offloading: Fine-Tuning LLMs with 100 Billion Parameters on a Single GPU

Faithful Vision-Language Interpretation via Concept Bottleneck Models

Towards Safe Concept Transfer of Multi-Modal Diffusion via Causal Representation Editing