I am a Ph.D. student in Computer Science at Brigham Young University (BYU), co-advised by Prof. Taylor Killian and Prof. Amanda Hughes. I received my M.S. in Computer Science and Engineering from the University of Notre Dame, and my B.Eng. in Computer Science from Beijing Jiaotong University.

My research is in AI/ML, with a focus on Large Language Models (LLMs), Vision-Language and Multimodal Large Language Models (VLMs/MLLMs), computer vision, and trustworthy AI. I am particularly interested in post-training and reasoning distillation, test-time scaling, multimodal hallucination mitigation, and image forensics. I work day-to-day with PyTorch, Hugging Face Transformers, vLLM, DeepSpeed, and PEFT/LoRA, building reproducible distributed training and evaluation pipelines.

I am actively seeking research internships in LLMs, VLMs, multimodal reasoning, generative AI, and AI safety. I am always happy to connect and discuss related ideas.

🔥 News

• 2026.02:  🎉 FRAME was accepted to the CVPR 2026 SAFE Workshop.
• 2026.01:  🎉 ICPO was accepted to ICLR 2026.

📝 Publications

(* denotes equal contribution)

Image Forensics & Multimodal AI Safety

• FRAME: Forensic Routing and Adaptive Multi-path Evidence Fusion for Image Manipulation Detection
  Kaixiang Zhao, Tianrun Yu, Aoxu Zhang, Junhao Su, Porter Jenkins, Amanda Hughes
  CVPR 2026 SAFE Workshop

• TIGER: Traceable Inference with Graph-Based Evidence Routing for Mitigating Hallucinations in Multimodal Generation
  Kaixiang Zhao, et al.
  EMNLP 2026 (under review)

LLM Reasoning, Distillation & Test-Time Optimization

• LARK: Learnability-Grounded Trajectory Selection for Efficient Reasoning Distillation
  Tianrun Yu, Kaixiang Zhao, et al.
  NeurIPS 2026 (under review)

• Provable and Practical In-Context Policy Optimization for Self-Improvement
  Tianrun Yu, Yuxiao Yang, Zhaoyang Wang, Kaixiang Zhao, Porter Jenkins, Xuchao Zhang, Chetan Bansal, Huaxiu Yao, Weitong Zhang
  ICLR 2026

Trustworthy AI & Model Security

• A Survey on Model Extraction Attacks and Defenses for Large Language Models
  Kaixiang Zhao, Lincan Li, Kaize Ding, Neil Zhenqiang Gong, Yue Zhao, Yushun Dong
  KDD 2025 Tutorial

• A Systematic Survey of Model Extraction Attacks and Defenses: State-of-the-Art and Perspectives
  Kaixiang Zhao, Lincan Li, Kaize Ding, Neil Zhenqiang Gong, Yue Zhao, Yushun Dong
  Preprint

• GraphIP-Bench: How Hard Is It to Steal a Graph Neural Network, and Can We Stop It?
  Kaixiang Zhao, Bolin Shen, Yuyang Dai, Shayok Chakraborty, Yushun Dong
  Preprint

• Intellectual Property in Graph-Based Machine Learning as a Service: Attacks and Defenses
  Lincan Li, Bolin Shen, Chenxi Zhao, Yuxiang Sun, Kaixiang Zhao, Shirui Pan, Yushun Dong
  Preprint

Federated Learning

• When the Server Steps In: Calibrated Updates for Fair Federated Learning
  Tianrun Yu, Kaixiang Zhao, Cheng Zhang, Anjun Gao, Yueyang Quan, Zhuqing Liu, Minghong Fang
  WiOpt 2026

📖 Educations

• 2025.08 - 2028.05 (Expected), Ph.D. in Computer Science, Brigham Young University, Provo, USA
• 2024.08 - 2025.08, M.S. in Computer Science and Engineering, University of Notre Dame, USA
• 2020.09 - 2024.06, B.Eng. in Computer Science, Beijing Jiaotong University, Beijing, China

🛠 Skills

• Programming & Systems: Python, C/C++, Bash, SQL, Git, Linux, CUDA, Docker, Slurm, LaTeX
• ML/DL Frameworks: PyTorch, Hugging Face Transformers, JAX, TensorFlow, vLLM, DeepSpeed, Accelerate, PEFT/LoRA, FlashAttention, NumPy, Pandas, scikit-learn
• LLMs & Post-Training: supervised fine-tuning, instruction tuning, reasoning distillation, chain-of-thought reasoning, test-time scaling, policy/preference optimization, model evaluation
• VLMs, Vision & Multimodal AI: Vision-Language Models, Multimodal LLMs, computer vision, generative AI, image/audio-text modeling, image forensics, cross-modal consistency, hallucination mitigation
• Research Engineering: distributed training, multi-GPU inference, model serving, retrieval-augmented generation, embeddings & vector search, reproducible evaluation pipelines, benchmark design, ablation studies