Hai Duong

Ph.D. Student
George Mason University
hduong22@gmu.edu

About Me

I am a Ph.D. student in Computer Science at George Mason University. I am working with Prof. ThanhVu Nguyen at ROARS Lab on Deep Neural Network Verification.

Research Interests

DNN Verification and Testing
Formal Method, Automated Reasoning
Proof Generation and Checking

News

10/2025: Received NeurIPS’25 Scholar Award.
09/2025: NeurIPS’25 paper on generating and checking DNN verification proofs.
09/2025: NeurIPS’25 paper on compositional DNN verification (Spotlight).
08/2025: NeuralSAT ranked 2nd overall in VNN-COMP’25.
04/2025: CAV’25 paper on NeuralSAT verification tool.
07/2024: NeuralSAT ranked 2nd overall in VNN-COMP’24. GMU Press.
02/2024: FSE’24 paper on new optimizations developed for the NeuralSAT.
12/2023: TACAS’24 paper on a verifiable-friendly training method.
07/2023: NeuralSAT ranked 4th overall and received New Participant Award in VNN-COMP’23.

Publications

1. Compositional Neural Network Verification via Assume-Guarantee Reasoning

Hai Duong; David Shriver; ThanhVu Nguyen; Matthew Dwyer

Neural Information Processing Systems (NeurIPS'25) - Spotlight

PDF Code
2. Generating and Checking DNN Verification Proofs

Hai Duong; ThanhVu Nguyen; Matthew Dwyer

Neural Information Processing Systems (NeurIPS'25)

PDF Code
3. NeuralSAT: A High-Performance Verification Tool for Deep Neural Networks

Hai Duong; ThanhVu Nguyen; Matthew Dwyer

Computer Aided Verification (CAV'25)

PDF Code
4. NeuralSAT: Scaling Constraint Solving for DNN Verification

Hai Duong; ThanhVu Nguyen

International Symposium on AI Verification (SAIV'25) - Competition Contribution

PDF Code
5. Harnessing Neuron Stability to Improve DNN Verification

Hai Duong; Dong Xu; ThanhVu Nguyen; Matthew Dwyer

Foundations of Software Engineering (FSE'24)

PDF Code
6. Training for Verification: Increasing Neuron Stability to Scale DNN Verification

Dong Xu; Nusrat Mozumder; Hai Duong; Matthew Dwyer

Tools and Algorithms for the Construction and Analysis of Systems (TACAS'24)

PDF Code
7. SymInfer: Inferring Numerical Invariants using Symbolic States

ThanhVu Nguyen; KimHao Nguyen; Hai Duong

International Conference on Software Engineering (ICSE'22) - Companion Proceedings

PDF Code

Misc

NeurIPS Scholar Award, 2025
Contributed to proposal NVIDIA Academic Grant Program: Trustworthy AI: Bringing Scalability and Assurance to DNN Verification. 2025
Contributed to proposal NSF CAREER. NeuralSAT: A Constraint-Solving Framework for Verifying Deep Neural Networks. 2023
Contributed to proposal Amazon Research Award. Scalable and Precise DNN Constraint Solving with Abstraction and Conflict Clause Learning. 2023
Helped review papers: ISSTA’23, PLDI’24, OOPSLA’24, ICLR’25