Saikat Dutta
Saikat Dutta
I am an Assistant Professor in the Department of Computer Science at Cornell University. My research interests are at the intersection of Software Engineering and Machine Learning.

I received my PhD in Computer Science from the University of Illinois Urbana-Champaign in Summer 2023, advised by Prof. Sasa Misailovic. Before joining Cornell, I spent a year as a Postdoctoral Researcher at the University of Pennsylvania, working with Prof. Mayur Naik. You can find my CV here.

Prospective students: I am looking for skilled and motivated undergraduates, PhD students, and postdocs to join my group. If you are interested in working with me, please drop me an email. If you are a prospective PhD student, apply to the Cornell CS PhD program.

Research Interests
My research interests are at the intersection of Software Engineering and Machine Learning. I am particularly interested in 1) developing novel techniques and tools to improve the reliability of Machine Learning-based systems, and 2) leveraging Machine Learning to solve challenging tasks in Software Engineering.

My research focuses on following themes: Apart from these topics, I also developed novel inference algorithms and robustness analyses for probabilistic programming [UAI 2023] [ATVA 2021].
Awards
Mavis Future Faculty Fellowship 2022-23
Facebook PhD Fellowship 2020-22
3M Foundation Fellowship 2019-2020
News
  • New:
    Passed PhD Final Defense! Find my dissertation here.
  • New:
    Our paper ASTRA: Understanding the Practical Impact of Robustness for Probabilistic Programs has been accepted to UAI 2023!
  • New:
    Our paper Balancing Effectiveness and Flakiness of Non-Deterministic Machine Learning Tests has been accepted to ICSE 2023!
  • Our paper To Seed or Not to Seed? An Empirical Analysis of Usage of Seeds for Testing in Machine Learning Projects has been accepted to ICST 2022!
  • Our paper InspectJS: Leveraging Code Similarity and User-Feedback for Effective Taint Specification Inference for JavaScript has been accepted to ICSE-SEIP 2022!
    • More News
  • Our paper SixthSense: Debugging Convergence Problems in Probabilistic Programs via Program Representation Learning has been accepted to FASE 2022!
  • Our paper AQUA: Automated Quantized Inference for Probabilistic Programs has been accepted to ATVA 2021!
  • Our paper TERA: Optimizing Stochastic Regression Tests in Machine Learning Projects has been accepted to ISSTA 2021!
  • Our paper FLEX: Fixing Flaky Tests in Machine-Learning Projects by Updating Assertion Bounds has been accepted to ESEC/FSE 2021!
  • I will be interning at Amazon Web Services (AWS) with the Automated Reasoning Group (ARG) for Summer 2021!
  • Our paper on Detecting Flaky Tests in Probabilistic and Machine Learning Applications was accepted to ISSTA 2020!
  • I will be interning at Microsoft Research, Redmond with the RISE group for Summer 2020!
    Looking forward to it!
  • Awarded Facebook PhD Fellowship 2020
    Thanks Facebook!
  • Our paper, Storm: Program Reduction for Testing and Debugging Probabilistic Programming Systems, has been accepted to FSE 2019
  • Selected for 3M Foundation Fellowship 2018-19
  • Our paper on ProbFuzz, "Testing Probabilistic programming systems" has been accepted to FSE 2018
  • Attended PLDI 2018 at Philadelphia, USA (20-22 June, 2018)
  • Our recent work on Automated Sensitivity Analysis was published in IEEE TSE Volume 43, Issue 12
  • Attended Midwest Programming Language Summit 2017 at Bloomington, Indiana
  • Attended Automated Software Engineering Conference (ASE 2017) at UIUC
Selected Publications
Balancing Effectiveness and Flakiness of Non-Deterministic Machine Learning Tests
45th International Conference on Software Engineering (ICSE 2023)
Steven Xia, Saikat Dutta, Sasa Misailovic, Darko Marinov, and Lingming Zhang
[Paper] [Code]
FLEX: Fixing Flaky Tests in Machine-Learning Projects by Updating Assertion Bounds
29th ACM Joint European Software Engineering Conference and Symposium on the Foundations of Software Engineering (FSE 2021)
Saikat Dutta, August Shi, and Sasa Misailovic
[Paper] [Code]
Testing Probabilistic Programming Systems
26th ACM Joint European Software Engineering Conference and Symposium on the Foundations of Software Engineering (FSE 2018)
Saikat Dutta, Owolabi Legunsen, Zixin Huang, Sasa Misailovic
[Paper] [Code]
All Publications
2024
LLM-Assisted Static Analysis for Detecting Security Vulnerabilities
Ziyang Li, Saikat Dutta, Mayur Naik
May 2024
[Preprint]
Understanding the Effectiveness of Large Language Models in Detecting Security Vulnerabilities
Avishree Khare, Saikat Dutta, Ziyang Li, Alaia Solko-Breslin, Rajeev Alur, Mayur Naik
June 2024
[Preprint]
Debugging Convergence Problems in Probabilistic Programs via Program Representation Learning with SixthSense
The International Journal on Software Tools for Technology Transfer (STTT 2024)
Zixin Huang, Saikat Dutta, and Sasa Misailovic
Extended version of the FASE 2022 paper
2023
Randomness-Aware Testing of Machine Learning-based Systems
Ph.D. Dissertation, University of Illinois Urbana-Champaign, July 2023
Saikat Dutta
[Dissertation]
ASTRA: Understanding the Practical Impact of Robustness for Probabilistic Programs
39th Conference on Uncertainty in Artificial Intelligence (UAI 2023)
Pittsburgh, PA, August 2023. Acceptance Rate 31% (243/778 papers)
Zixin Huang, Saikat Dutta, and Sasa Misailovic
[Paper]
Balancing Effectiveness and Flakiness of Non-Deterministic Machine Learning Tests
45th International Conference on Software Engineering (ICSE 2023)
Melbourne, Australia, May 2023. Acceptance Rate 26% (208/796 papers)
Steven Xia, Saikat Dutta, Sasa Misailovic, Darko Marinov, and Lingming Zhang
[Paper] [Code]
2022
To Seed or Not to Seed? An Empirical Analysis of Usage of Seeds for Testing in Machine Learning Projects
15th IEEE International Conference on Software Testing, Verification and Validation (ICST 2022)
Valencia, Spain, April 2022. Acceptance Rate 29% (25/85 papers)
Saikat Dutta, Anshul Arunachalam and Sasa Misailovic
[Paper] [Code]
InspectJS: Leveraging Code Similarity and User-Feedback for Effective Taint Specification Inference for JavaScript
44th International Conference on Software Engineering - Software Engineering in Practice (ICSE-SEIP 2022)
Pittsburgh, USA, May 2022.
Saikat Dutta, Diego Garbervetsky, Shuvendu Lahiri, Max Schäfer
[Paper] [Slides]
SixthSense: Debugging Convergence Problems in Probabilistic Programs via Program Representation Learning
25th International Conference on Fundamental Approaches to Software Engineering (FASE 2022)
Munich, Germany, April 2022. Acceptance Rate 27% (17/62 papers)
Saikat Dutta, Zixin Huang, and Sasa Misailovic
[Paper] [Code]
2021
Automated Quantized Inference for Probabilistic Programs with AQUA
Innovations in Systems and Software Engineering: A NASA Journal (ISSE NASA)
Zixin Huang, Saikat Dutta, and Sasa Misailovic
[Paper]
Extended version of our ATVA 2021 paper
AQUA: Automated Quantized Inference for Probabilistic Programs
19th International Symposium on Automated Technology for Verification and Analysis (ATVA 2021)
Gold Coast, Australia, October 2021. Acceptance Rate 27% (19/71 papers)
Zixin Huang, Saikat Dutta, and Sasa Misailovic
[Paper] [Code]
FLEX: Fixing Flaky Tests in Machine-Learning Projects by Updating Assertion Bounds
29th ACM Joint European Software Engineering Conference and Symposium on the Foundations of Software Engineering (FSE 2021)
Athens, Greece, August 2021. Acceptance rate 24% (97/396 papers)
Saikat Dutta, August Shi, and Sasa Misailovic
[Paper] [Code]
TERA: Optimizing Stochastic Regression Tests in Machine Learning Projects
30th ACM SIGSOFT International Symposium on Software Testing and Analysis (ISSTA 2021)
Aarhus, Denmark, July 2021. Acceptance rate 22% (51/233 papers)
Saikat Dutta, Jeeva Selvam, Aryaman Jain, and Sasa Misailovic
[Paper] [Code]
2020
Detecting Flaky Tests in Probabilistic and Machine Learning Applications
29th ACM SIGSOFT International Symposium on Software Testing and Analysis (ISSTA 2020)
Los Angeles, CA, USA, July 2020. Acceptance rate 26% (43/162 papers)
Saikat Dutta, August Shi, Rutvik Choudhary, Zhekun Zhang, Aryaman Jain, and Sasa Misailovic
[Paper] [Code]
2019
Storm: Program Reduction for Testing and Debugging Probabilistic Programming Systems
27th ACM Joint European Software Engineering Conference and Symposium on the Foundations of Software Engineering (ESEC/FSE 2019)
Talin, Estonia, August 2019. Acceptance rate 24% (74/303 papers)
Saikat Dutta, Wenxian Zhang, Zixin Huang, Sasa Misailovic
[Paper] [Code]
2018
Testing Probabilistic Programming Systems
26th ACM Joint European Software Engineering Conference and Symposium on the Foundations of Software Engineering (ESEC/FSE 2018)
Lake Buena Vista, FL, USA, November 2018. Acceptance rate 21% (61/289 papers)
Saikat Dutta, Owolabi Legunsen, Zixin Huang, Sasa Misailovic
[Paper] [Code]
2013-17
AutoSense: A Framework for Automated Sensitivity Analysis of Program Data
IEEE Transactions on Software Engineering (TSE 2017)
Bernard Nongpoh, Rajarshi Ray, Saikat Dutta, Ansuman Banerjee
[Paper]
Enhancing branch prediction using software evolution
10th IEEE International Conference on Networking, Architecture, and Storage (NAS 2015)
Saikat Dutta, Moumita Das, Ansuman Banerjee
A New Approach for Minimal Environment Construction for Modular Property Verification
24th Asian Test Symposium (ATS 2015)
Saikat Dutta, Soumi Chattopadhyay, Ansuman Banerjee, Pallab Dasgupta
A Framework for Fast Service Verification and Query Execution for Boolean Service Rules>
9th Asia-Pacific Services Computing Conference (APSCC 2015)
Soumi Chattopadhyay, Saikat Dutta, Ansuman Banerjee
Daikon to Prioritize and Group Unit Bugs
Formal Aspects of Component Software - 10th International Symposium (FACS 2013)
Nehul Jain, Saikat Dutta, Ansuman Banerjee, Anil K. Ghosh, Lihua Xu, Huibiao Zhu
Service
ISSTA 2025 Program Committee
ASE 2024 Program Committee
MLSYS 2024 Program Committee
TSE 2022 Reviewer
MSR Shadow PC 2022
PLDI 2021 Artifact Evaluation Committee
OOPSLA 2020 Artifact Evaluation Committee