About Me

Mission-driven undergraduate student researcher at Stony Brook University, motivated to pursue challenging, high-impact problems across a rich spectrum of AI topics. I have been previously guided by Dr. Keze Wang during two formative years at Sun Yat-sen University's HCP Lab—a foundation that keeps me exploring widely and actively seeking new collaborations.

Education

Stony Brook University, Stony Brook, NY — B.S., Applied Mathematics and Statistics (AMS); Computer Science AOI

Experience

Microsoft Research Asia (MSRA)
Beijing, China  ·  Research Intern

Working on RD-Agent GitHub stars — an open-source AI-driven framework that automates data-driven R&D processes, enabling AI to drive data-driven AI.

Projects

An evidence-backed skills library with 101,000+ curated skills across 21 domains, extracted from academic papers (PMC, arXiv, eLife) and technical sources (GitHub, StackOverflow). Features full-text search over offline SQLite bundles and an automated LLM-powered generation pipeline with source traceability.

Publications

Under Review / Submitted

  • Cheng, Z., Li, J., Gao, Y., Sha, T., Tang, Y., Chang, L.-L., Dong, Y. WeatherBench-R: A Text-Only Benchmark for Evaluating Large Language Models over U.S. Weather Events. ACL ARR 2026
  • Cheng, Z., Xu, H., Sha, T., Zhao, Y., Li, M., Dong, Y. TRIDENT: An Efficient Data-Free Model Extraction Attack for Graph Neural Networks. ICLR 2026
  • Chen, Z., Sha, T., Chen, Y., Xu, B., Zheng, Y., Cheng, Z., Liu, J., Wang, K. DIAL-G²: Graph-Guided Dialectical Agent for Advanced ESG Reasoning. ICLR 2026

Accepted & Published

  • Sha, T., Cheng, Z., Zhai, H., Ding, X., Chen, Z., Wang, K. FAST-CAD: Fairness-Aware Self-Supervised Framework for Automated Non-Contact Stroke Diagnosis. AAAI 2026 Oral
  • Cheng, Z., Shen, B., Sha, T., Gao, Y., Dong, Y. ATOM: A Framework of Detecting Query-Based Model Extraction Attacks for Graph Neural Networks. KDD 2025
  • Zhai, H., Chen, X., Zhang, C., Sha, T., Li, R. Mitigating Cache Noise in Test-Time Adaptation for Large Vision-Language Models. ICME 2025 Oral ICLR 2025 Workshop
  • Chen, Z., Sha, T., Tang, Z., Wang, K. TsKAN: A Transparent Architecture for Improving the Interpretability of Multivariate Time Series Forecasting. ICLR 2025 Workshop
  • Liao, J., Sha, T., Wang, Q., Zhong, Z., Lin, H. Orthogonal Filtering Alignment Network for Ship Detection in SAR Images Under Frequency Shift Interference. IGARSS 2025

Awards and Honors

  • AAAI-26 Student Scholarship & Volunteer Program — Travel scholarship and selected student volunteer.
  • China National Olympiad in Informatics (NOI) Winter Camp 2022 — Silver Medal (roughly comparable to USACO Platinum).
  • NOIP 2021 — Provincial First Prize (Guangdong) (roughly comparable to USACO Gold).
  • National College Student Mathematical Modeling Competition 2024 — Team Leader/Member, Provincial First Prize.

Service

Conference Reviewer

AAAI 2026 ICLR 2025 ICML 2025 KDD 2025 CVPR 2026 ICME 2026

Journal Reviewer

IEEE TNNLS Neural Networks

Interview

MIT Technology Review ChinaDeveloping a non-contact AI tool to predict stroke with >91% accuracyRead more (Nov 15, 2025)