AI Coding Researcher
ByteDance TRAE Research
Email:
Personal Email:
I am currently a researcher at ByteDance TRAE Research, focusing on AI Coding research. I received my Ph.D. degree in Computer Science from ShanghaiTech University in 2023, under the supervision of Prof. Fu Song. During my Ph.D. studies, I was fortunate to be a visiting student at Singapore Management University, working with Prof. Jun Sun.
My current research interests are centered on AI Coding. This includes designing and evaluating Coding Agents from a software engineering (or formal verification) perspective, exploring the characteristics of LLMs on coding tasks to push their capability boundaries, token-saving strategies for Coding Agents, and the training of Code LLMs.
[20] Trae Agent: An LLM-based Agent for Software Engineering with Test-time Scaling
arXiv:2507.23370, 2025
#1 on SWE-bench Verified 🏆
[19] CodeVisionary: An Agent-based Framework for Evaluating Large Language Models in Code Generation
ASE 2025
[18] Improving the Efficiency of LLM Agent Systems through Trajectory Reduction
arXiv:2509.23586, 2025
[17] Tool-integrated Reinforcement Learning for Repo Deep Search
arXiv:2508.03012, 2025
[16] SoRFT: Issue Resolving with Subtask-oriented Reinforced Fine-Tuning
ACL 2025
[15] Verification of Bit-Flip Attacks against Quantized Neural Networks
OOPSLA 2025
[14] AEGIS: An Agent-based Framework for General Bug Reproduction from Issue Descriptions
FSE 2025 Industry
[13] An Empirical Study on LLM-based Agents for Automated Bug Fixing
arXiv:2411.10213, 2024
[12] MarsCode Agent: AI-native Automated Bug Fixing
arXiv:2409.00899, 2024
[11] RepoMasterEval: Evaluating Code Completion via Real-World Repositories
ASE Industry Showcase 2025 (accepted)
[10] SAT-based Formal Verification of Fault Injection Countermeasures for Cryptographic Circuits
IACR Transactions on Cryptographic Hardware and Embedded Systems, 2024(4)
[9] Compositional Verification of First-Order Masking Countermeasures against Power Side-Channel Attacks
ACM Transactions on Software Engineering and Methodology, 33(3): 79:1-79:38, 2024
[8] Compositional Verification of Efficient Masking Countermeasures against Side-Channel Attacks
Proceedings of the ACM on Programming Languages, 7(OOPSLA2): 286:1817–286:1847, 2023
[7] VenomAttack: Automated and Adaptive Activity Hijacking in Android
Frontiers of Computer Science, 17(1):171801, 1-18, 2023
[6] Model-based Automated Testing of JavaScript Web Applications via Longer Test Sequences
Frontiers of Computer Science, 16(3):163204, 1-14, 2022
[5] Formal Verification of Masking Countermeasures for Arithmetic Programs
IEEE Transactions on Software Engineering, 48(3): 973-1000, 2022
[4] A Hybrid Approach to Formal Verification of Higher-Order Masked Arithmetic Programs
ACM Transactions on Software Engineering and Methodology, 30(3):1-42, 2021
[3] Verifying and Quantifying Side-Channel Resistance of Masked Software Implementations
ACM Transactions on Software Engineering and Methodology, 28(3):16:1-16:32, 2019
[2] Quantitative Verification of Masked Arithmetic Programs against Side-Channel Attacks
Proceedings of the 25th International Conference on Tools and Algorithms for the Construction and Analysis of Systems (ETAPS/TACAS), Prague, Czech Republic. April 6-11, 2019
[1] SCInfer: Refinement-based Verification of Software Countermeasures against Side-Channel Attacks
Proceedings of the 30th International Conference on Computer Aided Verification (CAV), Oxford, UK. July 14-17, 2018