Interpreting Robustness Proofs of Deep Neural Networks
Debangshu Banerjee, Avaljot Singh, Gagandeep Singh, Arxiv 2023.
CV
Education
- Ph.D. in Computer Science, ETH Zurich, 2020
- Masters in Computer Science, ETH Zurich, 2014
- ETH Medal for Outstanding Master Thesis
- ETH Excellence Scholarship
- B.Tech. in Computer Science and Engineering, IIT Patna, 2012
- President of India Gold Medal
- Institute Silver Medal
Work experience
- Nov 2020-Aug 2021: Researcher, VMware Research
- Summer 2011: Research Assistant, University of New South Wales
- Summer 2010: Research Assistant, University of Houston
Publications
CoMEt: x86 Cost Model Explanation Framework
Isha Chaudhary, Alex Renda, Charith Mendis, Gagandeep Singh, Arxiv 2023.
Incremental Verification of Neural Networks
Shubham Ugare, Debangshu Banerjee, Sasa Misailovic, and Gagandeep Singh, PLDI 2023.
Black-Box Targeted Reward Poisoning Attack Against Online Deep Reinforcement Learning
Yinglun Xu, Gagandeep Singh, Arxiv 2023.
Provable Defense Against Geometric Transformations
Rem Yang, Jacob Laurel, Sasa Misailovic, Gagandeep Singh, ICLR 2023 (Spotlight).
Exploring Practical Vulnerabilities of Machine Learning-based Wireless Systems
Zikun Liu, Changming Xu, Gagandeep Singh, and Deepak Vasisht, NSDI 2023.
Property-Driven Evaluation of RL-Controllers in Self-Driving Datacenters
Arnav Chakravarthy, Nina Narodytska, Asmitha Rathis, Marius Vilcu, Mahmood Sharif, Gagandeep Singh, DMML@NeurIPS 2022.
Physically-Constrained Adversarial Attacks on Brain-Machine Interfaces
Xiaying Wang, Rodolfo Octavio Siller Quintanilla, Michael Hersche, Luca Benini, Gagandeep Singh, TSRML@NeurIPS 2022.
Scalable Verification of GNN-Based Job Schedulers
Haoze Wu, Clark Barrett, Mahmood Sharif, Nina Narodytska, Gagandeep Singh, OOPSLA 2022.
Proof Transfer for Fast Certification of Multiple Approximate Neural Networks
Shubham Ugare, Gagandeep Singh, Sasa Misailovic, OOPSLA 2022.
A General Construction for Abstract Interpretation of Higher-Order Automatic Differentiation
Jacob Laurel, Rem yang, Shubham Ugare, Robert Nagel, Gagandeep Singh, Sasa Misailovic, OOPSLA 2022.
Shared Certificates for Neural Network Verification
Marc Fischer, Christian Sprecher, Dimitar I. Dimitrov, Gagandeep Singh, Martin Vechev, CAV 2022.
Robust Universal Adversarial Perturbations
Changming Xu, Gagandeep Singh, Arxiv 2022.
Efficient Reward Poisoning Attacks on Online Deep Reinforcement Learning
Yinglun Xu, Qi Zeng, Gagandeep Singh, Arxiv 2022.
Provably Robust Adversarial Examples
Dimitar I. Dimitrov, Gagandeep Singh, Timon Gehr, Martin Vechev, ICLR 2022.
PRIMA: General and Precise Neural Network Certification via Scalable Convex Hull Approximations
Mark Niklas Müller, Gleb Makarchuk, Gagandeep Singh, Markus Püschel, Martin Vechev, POPL 2022.
A Dual Number Abstraction for Static Analysis of Clarke Jacobians
Jacob Laurel, Rem Yang, Gagandeep Singh, Sasa Misailovic, POPL 2022.
FIRE: Enabling Reciprocity for FDD MIMO Systems
Zikun Liu, Gagandeep Singh, Chenren Xu, Deepak Vasisht, MobiCom 2021.
Robustness Certification for Point Cloud Models
Tobias Lorenz, Anian Ruoss, Mislav Balunović, Gagandeep Singh, Martin Vechev, ICCV 2021.
Scalable Polyhedral Verification of Recurrent Neural Networks
Wonryong Ryou, Jiayu Chen, Mislav Balunovic, Gagandeep Singh, Andrei Dan, Martin Vechev, CAV 2021.
Scaling Polyhedral Neural Network Verification on GPUs
Christoph Müller, Francois Serre, Gagandeep Singh, Markus Püschel, Martin Vechev, MLSys 2021.
Adversarial Attacks on Probabilistic Autoregressive Forecasting Models
Raphaël Dang-Nhu, Gagandeep Singh, Pavol Bielik, Martin Vechev, ICML 2020.
Learning Fast and Precise Numerical Analysis
Jingxuan He, Gagandeep Singh, Markus Püschel, Martin Vechev, PLDI 2020.
Beyond the Single Neuron Convex Barrier for Neural Network Certification
Gagandeep Singh, Rupanshu Ganvir, Markus Püschel, Martin Vechev, NeurIPS 2019.
Certifying Geometric Robustness of Neural Networks
Mislav Balunovic, Maximilian Baader, Gagandeep Singh, Timon Gehr, Martin Vechev, NeurIPS 2019.
Boosting Robustness Certification of Neural Networks
Gagandeep Singh, Timon Gehr, Markus Püschel, Martin Vechev, ICLR 2019.
An Abstract Domain for Certifying Neural Networks
Gagandeep Singh, Timon Gehr, Markus Püschel, Martin Vechev, POPL 2019.
Fast and Effective Robustness Certification
Gagandeep Singh, Timon Gehr, Matthew Mirman, Markus Püschel, Martin Vechev, NeurIPS 2018.
Fast Numerical Program Analysis with Reinforcement Learning
Gagandeep Singh, Markus Püschel, Martin Vechev, CAV 2018.
A Practical Construction for Decomposing Numerical Abstract Domains
Gagandeep Singh, Markus Püschel, Martin Vechev, POPL 2018.
Fast Polyhedra Abstract Domain
Gagandeep Singh, Markus Püschel, Martin Vechev, POPL 2017.
Making Numerical Program Analysis Fast
Gagandeep Singh, Markus Püschel, Martin Vechev, PLDI 2015.
Past Students
Makarchuk Gleb, Dimitar Dimitrov, Clemens Giuliani, Raphaël Dang-Nhu, Jonathan Mauer, Rupanshu Ganvir, Christoph Müller, Jovan Andonov, Jakub Kotal, Afra Amini
Talks
Certified Artificial Intelligence
Invited Tutorial at WASP summer school in AI, Virtual
Certified Artificial Intelligence
Invited Talk at VMware Research, Tel Aviv University, UIUC, Cornell, Rice, NYU, MIT, and Georgia Tech,
Safe and Robust Deep Learning
Invited Talk at Workshop on Dependable and Secure Software Systems, ETH Zurich, Switzerland
Safe and Robust Deep Learning
Invited Talk at Waterloo ML + Security + Verification Workshop, University of Waterloo and New York University,
Safe and Robust Deep Learning
Invited Tutorial at DigiCosme Spring School on Formal Methods and Machine Learning, ENS Paris-Saclay, Cachan, France
An Abstract Domain for Certifying Neural Networks
Invited Talk at ACM Principles of Programming Languages (POPL), Lisbon, Portugal
Fast Numerical Program Analysis with Reinforcement Learning
Invited Talk at Computer Aided Verification (CAV), Oxford, UK
A Practical Construction for Decomposing Numerical Abstract Domains
Invited Talk at ACM Principles of Programming Languages (POPL), Los Angeles, USA
Fast Polyhedra Abstract Domain
Invited Talk at ACM Principles of Programming Languages (POPL), Paris, France
Making Numerical Program Analysis Fast
Invited Talk at ACM Programming Language Design and Implementation (PLDI), Portland, USA
Software
Teaching
Trustworthy AI Systems
Logic and Artificial Intelligence
Advanced Systems Lab
Reliable and Interpretable Artificial Intelligence
Blockchain Security Seminar
Research Topics in Software Engineering
How to Write Fast Numerical Code
Software Engineering Seminar
Program Analysis for System Security and Reliability
Algorithms and Data Structures
Program Analysis and Synthesis
Service
ICSE’24, PLDI’21,’23, ICLR’22,’23, AAAI’21,’22,’23, MLSys’23, SAS’21, ‘22 (Co-Chair), NeurIPS’,’19,’21,’22, ICML’22, CCS’21,’22, CVPR’21,’22, FOMLAS’20,’21,’22, ESOP’22, VMCAI’22, APLAS’21, ICCV’21, ATVA’21, IJCAI’21, TACAS’21, ASPLOS’21, JMLR, TMLR, FAOC, Machine Learning, IEEE TNNLS, TOPLAS