Xiaodong Yang
Staff Research Scientist
Research areas: Adversarial robustness of LLM-based agentic systems, self-evolving agent harness, LLM fine-tuning, graph neural networks
Dr. Xiaodong Yang joined Visa Research as a Staff Research Scientist in July 2022. Xiaodong received his Ph.D. in Computer Science from Vanderbilt University in 2022, where his research focused on verification, safety and robustness of deep neural networks and learning-enabled systems.
As a member of the Trustworthy AI team, Xiaodong’s research interests are broadly in trustworthy large language models, graph learning, recommendation systems and efficient machine learning. Recently, his research has focused on adversarial robustness of LLMs, self-evolving agent harnesses and LLM fine-tuning. He has also worked on agentic commerce systems, developing automated approaches to identify jailbreak and prompt injection strategies, as well as large-scale LLMs for transaction analysis and fraud detection. Xiaodong is passionate about developing secure, reliable and efficient AI systems and applying them to real-world financial applications.
Publications
- Zhang, Yilang, Xiaodong Yang, Yiwei Cai, and Georgios B. Giannakis. "ScaLoRA: Optimally Scaled Low-Rank Adaptation for Efficient High-Rank Fine-Tuning." ICML. 2026. “ScaLoRA: Optimally Scaled Low-Rank Adaptation for Efficient High-Rank Fine-Tuning (arxiv.org”
- Wang, Song, Xiaodong Yang, Rashidul Islam, Huiyuan Chen, Minghua Xu, Jundong Li, and Yiwei Cai. "Enhancing distribution and label consistency for graph out-of-distribution generalization." In 2024 IEEE international conference on data mining (ICDM). IEEE, 2024. “Enhancing Distribution and Label Consistency for Graph Out-of-Distribution Generalization”
- Yang, Xiaodong, Xiaoting Li, Huiyuan Chen, and Yiwei Cai. "Gaim: Attacking graph neural networks via adversarial influence maximization." 2025 SIAM International Conference on Data Mining (SDM). “GAIM: Attacking Graph Neural Networks via Adversarial Influence Maximization”
- Tran, Hoang Dung, Sung Woo Choi, Xiaodong Yang, Tomoya Yamaguchi, Bardh Hoxha, and Danil Prokhorov. "Verification of recurrent neural networks with star reachability." In Proceedings of the 26th ACM International Conference on Hybrid Systems: Computation and Control. 2023. “Verification of Recurrent Neural Networks with Star Reachability”
- Yang, Xiaodong, Tom Yamaguchi, Hoang-Dung Tran, Bardh Hoxha, Taylor T. Johnson, and Danil Prokhorov. "Neural network repair with reachability analysis." In International Conference on Formal Modeling and Analysis of Timed Systems. 2022. “Neural Network Repair with Reachability Analysis(http://arxiv.org )”
- Tran, Hoang-Dung, Xiaodong Yang, Diego Manzanas Lopez, Patrick Musau, Luan Viet Nguyen, Weiming Xiang, Stanley Bak, and Taylor T. Johnson. "NNV: the neural network verification tool for deep neural networks and learning-enabled cyber-physical systems." In International conference on computer aided verification, 2020.