What we are working on?
Our research aims to ensure the correctness of Deep Learning Systems, building more reliable, effective, and efficient AI systems for software engineering. Specifically, we focus on the following AI Infrastructures Testing:
- Deep Learning Library/Framework (PyTorch, Tensorflow, Mindspore,etc.)
- Deep Learning compilers (Inductor, TVM, ONNXRuntime,etc.)
- End-side deployment (JavaScript Engine, Autonomous Driving, etc.)
Team
Faculty
Chunrong Fang
Associate Professor fangchunrong@nju.edu.cnChunrong Fang received the B.E. and Ph.D. degrees in software engineering from Software Institute, Nanjing University, Jiangsu, China. He is currently an Associate Professor with the Software Institute of Nanjing University. His research interests lie in intelligent software engineering, e.g. BigCode and AITesting.
Zhenyu Chen
Professor zychen@nju.edu.cnZhenyu Chen is currently a full professor with Software Institute of Nanjing University. He is an associate Editor of IEEE Transactions on Reliability. He is also the Contest Co-Chair at QRS 2018, ICST 2019, and ISSTA 2019. He is the Industrial Track Co-Chair of SANER 2019. His research interests include collective intelligence, deep learning testing and optimization, big data quality, and mobile application testing.
Ph.D. Students
Jiawei Liu
Ph.D. Student, Fall 2021 jw.liu@smail.nju.edu.cn
Yanzhou Mu
Ph.D. Student, Fall 2022 602022320006@smail.nju.edu.cn
Yinglong Zou
Ph.D. Student, Fall 2023 652023320004@smail.nju.edu.cnMaster Students
Peiran Yang
Master Student, Fall 2023 peiranyang@smail.nju.edu.cn
Shaoyu Yang
Master Student, Fall 2024 shaoyuyoung@gmail.com
Yupeng Zhang
Master Student, Fall 2024 522024320213@smail.nju.edu.cnPublications
- All
- ASE
- ISSTA
- ICSE
- TOSEM
- TRel
- JSS
Bug List
* Currently, artifacts of bug finding method can be found in manuscript. In the future, artifacts would be open-source in our GitHub Organization
| System | #Issue Id | Symptom | Bug Finding Method | Status |
|---|---|---|---|---|
| TensorFlow.js | #8222 | crash (covert error) | DLJSFuzzer (ASE 2024) | Confirmed |
| TensorFlow.js | #7202 | xxxx | DLJSFuzzer (ASE 2024) | Pending |
| TensorFlow.js | #8246 | xxxx | DLJSFuzzer (ASE 2024) | Confirmed |
| TensorFlow.js | #8338 | xxxx | DLJSFuzzer (ASE 2024) | Confirmed |
| TensorFlow.js | #8337 | xxxx | DLJSFuzzer (ASE 2024) | Confirmed |
| TensorFlow.js | #8339 | xxxx | DLJSFuzzer (ASE 2024) | Confirmed |
| TensorFlow.js | #8340 | xxxx | DLJSFuzzer (ASE 2024) | Confirmed |
| TensorFlow.js | #8341 | xxxx | DLJSFuzzer (ASE 2024) | Confirmed |
| keras | #15666 | xxxx | Gandalf (TOSEM 2023) | Confirmed |
| keras | #15667 | xxxx | Gandalf (TOSEM 2023) | Confirmed |
| keras | #15677 | xxxx | Gandalf (TOSEM 2023) | Confirmed |
| keras | #15716 | xxxx | Gandalf (TOSEM 2023) | Confirmed |
| keras | #15717 | xxxx | Gandalf (TOSEM 2023) | Confirmed |
| TensorFlow | #53055 | xxxx | Gandalf (TOSEM 2023) | Confirmed |
| TensorFlow | #53107 | xxxx | Gandalf (TOSEM 2023) | Confirmed |
| keras | #68321 | xxxx | Gandalf (TOSEM 2023) | Confirmed |