Marathon

I am an undergraduate CS student at Dalian University of Technology. My research interests focus on Vision-Language-Action (VLA) and robot manipulation, with an emphasis on robustness and self-supervised learning under distribution shifts. I led work on improving zero-shot generalization for VLA models and contributed to Open-Vocabulary Pedestrian Attribute Recognition; both are currently under submission to ICML 2026. I also led the development of MindCare, a multimodal mental-health assessment system, and worked on robust multimodal forgery detection during an algorithm internship. I aim to develop reproducible, scalable methods that improve the reliability and generalization of VLA models in real-world settings.