Tiantai Zhang

 

Ph.D. Student, Nanjing University
Research Intern, Tencent Zhuque Lab (Qingyun Program)

Email: z472421519@gmail.com

Google Scholar  /  GitHub

[Now] [Research] [Publications] [Awards]

Bio

Tiantai Zhang is a Ph.D. student in the Department of Computer Science at Nanjing University, advised by Prof. Bing Mao (since September 2021). He is currently a research intern at Tencent Zhuque Lab through the Qingyun Program. He was previously a Visiting Scholar at Purdue University (June 2025–July 2026), working with Prof. Xiangyu Zhang.

His work spans reverse engineering, binary defense, and the application of large language models to binary security analysis. He has published at top venues including USENIX Security, ISSTA, and TOSEM. Before his Ph.D., he received his B.E. in Communication Engineering from Xi'an University of Posts and Telecommunications (2017–2021).

What I'm Working On

My current work asks how we can reason reliably about imperfect binaries and build practical protections whose security guarantees survive real-world compilation.

Audit & Validation
How can AI-assisted binary auditing move from plausible findings to verifiable security evidence?
Exploring how to move from plausible vulnerability findings to conclusions that can be checked against real program behavior.
Decompilation & AI
How do decompilation challenges affect model-based binary bug detection?
Using decompilation results to study which lost, distorted, or recovered program properties matter when models search binaries for bugs.
Defense & Optimization
Can compiler and runtime defenses remain secure and practical under optimization?
Investigating protections for security-sensitive program data while studying how compiler transformations affect their effectiveness, guarantees, and cost.

Updated July 2026

A white-hat researcher leads Claude, GPT, Kimi, DeepSeek, and GLM agents from source-code analysis across a semantic wall toward the binary-security frontier while defending against adversarial agents.
From source-code intelligence to the binary-security frontier.

Research Directions

Binary Analysis & Reverse Engineering
  • Binary disassembly ground truth and evaluation USENIX Security'22
  • Function entry identification and obfuscation ISSTA'23 TOSEM'25
  • Model-based binary bug detection under decompilation challenges Current
guided by Dr. Chengbin Pang where my research journey began
Binary Defense
  • Structure layout randomization Current
  • Compiler and runtime data protection under real-world optimization Current
introduced to this direction by Dr. Chengbin Pang my first step in broadening my research
LLM-Based Binary Security
  • Neural-symbolic vulnerability auditing for firmware binaries Current
  • Feedback-guided validation of binary vulnerability findings Current
introduced to this direction by Prof. Xiangyu Zhang and Chengpeng Wang my second step in that exploration

Publications

Awards — things that happened, things that mattered.

  • Nanjing University Graduate First-Class Academic Scholarship, 2022 & 2023

  • Nanjing University Ru Li Scholarship, 2022

  • ICPC Asia-East Continent Final, Bronze Medal, 2019

  • ICPC Asia Regional Contest, Nanjing Site, Bronze Medal, 2019

  • China Collegiate Programming Contest (CCPC), Jilin Site, Bronze Medal, 2018

  • ICPC Asia Regional Contest, Xuzhou Site, Silver Medal, 2018

Algorithm & Competitive Programming · team 🍬 I Love Candy w/ Wenfeng Huang & Chenlei Yang

Codeforces z472421519 1960 Candidate Master 1🥈 6🥉 4🏳️
🈚️ 🥉 🥉🥉 🥉🥉🥈 🥉🥉🥈🥉 🥉🥉🥈🥉🥉 🥉🥉🥈🥉🥉🥉 🥉🥉🥈🥉🥉🥉🥉 ICPC EC-Final Shanghai '17 ICPC Invite🍬 Xi'an '18 CCPC🍬 Jilin '18 ICPC🍬 Xuzhou '18 ICPC🍬 Beijing '18 CCPC Final🍬 Shenzhen '18 ICPC EC-Final🍬 Xi'an '18 ICPC Invite🍬 Xi'an '19 ICPC🍬 Nanjing '19 ICPC🍬 Shanghai '19 ICPC EC-Final🍬 2019

Last modified: 2026.04