Pudding 07/22/2013 - 12/23/2021
I will love and miss you forever, my girl, my baby, and my angel.

About Me

Hi, my within 6-hop neighbors! I am Ying Song, a fourth-year PhD candidate in Information Science at the University of Pittsburgh under the supervision of Dr. Balaji Palanisamy. Before that, I received my Master degree in Statistics at the National University of Singapore.

Self-Profiling: I used to be a wanderer with a microscope for observation. Whenever something intriguing struck, I would dissect it with logic and find joy in uncovering the hidden architecture of “how and why”. Those were the main sources of my “Aha” moments.

Now, my curiosity transcends mere witnessing. It drives me to be an active creator. My new mission is to define new problems that will shape the future and generate value for the social good.

Research Interests

“Know the Enemy”: How can we transcend the traditional “Catch Me If You Can” game and move beyond the temporary patches toward fundamental attack-immunity? How to translate the adversarial technical backbone into enforceable policy frameworks?

Keywords: Machine learning security, data privacy (mainly focus on graph learning), and AI auditing and governance.

“Digital Saddle”: How can we saddle up uncontrollable AI systems and steer them toward human-centric tracks? How to conquer the increased entropy of real-world contexts, such as distribution shifts, noises, misinformation, etc?

Keywords: Trustworthy AI: security and privacy (main focus), fairness, robustness, explainability, etc.

“Living Graphs and Associated Risks”: Graphs are the topological blueprints of the living and digital worlds, from molecule graphs and gene interactions to causal provenance in cyber forensics. How can we unleash the potential of these applications while identifying and mitigating their associated risks (e.g., poisoning drug discovery systems, hierarchical privacy leakage, and anti-forensics)?

Keywords: Graph Learning, Security and Privacy, and AI for Science (drug discovery, genetics, cyber forensics).

Publications

*denote the corresponding author.

[1] Song. Y.* & Palanisamy. B. (2026). Taipan: A Query-free Transfer-based Multiple Sensitive Attribute Inference Attack Solely from Publicly Released Graphs. Submitted to USENIX.

[2] Song. Y.* & Palanisamy. B. (2026). GraphToxin: Reconstructing Full Unlearned Graphs from Graph Unlearning. Submitted to USENIX.

[3] Song. Y.*, Singh. R. & Palanisamy. B. (2025). Krait: A Backdoor Attack Against Graph Prompt Tuning. In the 3rd IEEE Conference on Secure and Trustworthy Machine Learning (SaTML).

[4] Song. Y.* & Palanisamy. B. (2024). MAPPING: Debiasing Graph Neural Networks for Fair Node Classification with Limited Sensitive Information Leakage. In the World Wide Web J. 27 (6), 74.

[5] Song. Z., Li. W., Jin. X., Ying. J., Zhang. X., Song. Y., Li. H. & Fan. Q.(2022). Genetics, Leadership, and Well-Being: An Investigation with a Large-Scale GWAS. In the Proceedings of the National Academy of Sciences of the United States of America (PNAS).

[6] Ng. J., Hubner. S., Teow. J., Song. Y., Wang. Y., Kaur. A., Frese. M., Song. Z., Lee. L. & Moriguchi. T.(2021). Development of cultural inventory on Asian countries and exploratory approach to predict innovation. In the 81st Annual Meeting of the Academy of Management.

Selected Scholarships

• Teaching Assistant Scholarship, University of Pittsburgh, 2024-2026.
• Full Scholarship for the 3rd IEEE Conference on Secure and Trustworthy Machine Learning (SaTML), 2025.
• SCI Fellowship, University of Pittsburgh, 2022-2024.
• National Scholarship, Ministry of Education of the People’s Republic of China, 2018
• Meritorious Winner of the Interdisciplinary Contest in Modeling (MCM/ICM), COMAP, USA, 2018
• Second Prize of the Eastern Area of China in the First China (Hengqin) International University Quantitative Finance Competition, Worldwide, 2018
• Jiangsu Provincial Government Scholarship Program of Overseas Studies at UCLA, JESIE, 2017

Just move on fearlessly and embrace more possibilities.