Publications

On Definitions and Abstractions

  1. CCS’23
    Unforgeability in Stochastic Gradient Descent
    Teodora Baluta, Ivica Nikolić, Racchit Jain, Divesh Aggarwal, and Prateek Saxena
    In Proceedings of the 2023 ACM SIGSAC Conference on Computer and Communications Security (CCS), 2023
  2. CCS’22
    Membership Inference Attacks and Generalization: A Causal Perspective
    Teodora BalutaShiqi Shen, S Hitarth, Shruti Tople, and Prateek Saxena
    In Proceedings of the 2022 ACM SIGSAC Conference on Computer and Communications Security (CCS), 2022
  3. SafeGen’24
    Unlearning in- vs. out-of-distribution data in LLMs under gradient-based methods
    Teodora Baluta, Pascal Lamblin, Daniel Tarlow, Fabian Pedregosa, and Gintare Karolina Dziugaite
    In Neurips Safe Generative AI Workshop, 2024

On Statistical Verifiability

  1. ICSE’21
    Scalable quantitative verification for deep neural networks
    Teodora Baluta, Zheng Leong Chua, Kuldeep S Meel, and Prateek Saxena
    In 2021 IEEE/ACM 43rd International Conference on Software Engineering (ICSE), 2021
  2. CCS’19
    Quantitative verification of neural networks and its security applications
    In Proceedings of the 2019 ACM SIGSAC Conference on Computer and Communications Security (CCS), 2019

On Differential Privacy

  1. CCS’22
    LPGNet: Link Private Graph Networks for Node Classification
    Aashish KolluriTeodora Baluta, Bryan Hooi, and Prateek Saxena
    In Proceedings of the 2022 ACM SIGSAC Conference on Computer and Communications Security (CCS), 2022
  2. CCS’21
    Private hierarchical clustering in federated networks
    Aashish KolluriTeodora Baluta, and Prateek Saxena
    In Proceedings of the 2021 ACM SIGSAC Conference on Computer and Communications Security (CCS), 2021

On Learnability of Analysis / Synthesis Rules

  1. SAT’23
    Explaining SAT Solving Using Causal Reasoning
    Jiong Yang, Arijit Shaw, Teodora Baluta, Mate Soos, and Kuldeep S. Meel
    In International Conference on Theory and Applications of Satisfiability Testing (SAT), 2023
  2. OOPSLA’23
    User-Customizable Transpilation of Scripting Languages
    Bo WangAashish Kolluri, Ivica Nikolić, Teodora Baluta, and Prateek Saxena
    Proceedings of the ACM on Programming Languages (OOPSLA), 2023
  3. FSE’21
    SynGuar: guaranteeing generalization in programming by example
    Bo WangTeodora BalutaAashish Kolluri, and Prateek Saxena
    In Joint Meeting on European Software Engineering Conference and Symposium on the Foundations of Software Engineering (ESEC/FSE), 2021
  4. NDSS’19
    One Engine To Serve’em All: Inferring Taint Rules Without Architectural Semantics.
    Zheng Leong Chua, Yanhao Wang, Teodora BalutaPrateek Saxena, Zhenkai Liang, and Purui Su
    In Network and Distributed System Security (NDSS) Symposium, 2019