publications

publications by categories in reversed chronological order. generated by jekyll-scholar.

2025

  1. bluff.png
    Proof or Bluff? Evaluating LLMs on 2025 USA Math Olympiad
    Ivo Petrov, Jasper Dekoninck, Lyuben Baltadzhiev, Maria Drencheva, Kristian Minchev, Mislav Balunović, Nikola Jovanović, and Martin Vechev
    arXiv, 2025
    arXiv
  2. ICML
    mathconstruct.png
    MathConstruct: Challenging LLM Reasoning with Constructive Proofs
    Mislav Balunovic, Jasper Dekoninck, Nikola Jovanovic, Ivo Petrov, and Martin T. Vechev
    In ICML, 2025
    arXiv
  3. ICLR
    anonymizers.png
    Language Models are Advanced Anonymizers
    Robin Staab, Mark Vero, Mislav Balunovic, and Martin Vechev
    In ICLR, 2025
    arXiv

2024

  1. agentsformal.png
    AI Agents with Formal Security Guarantees
    Mislav Balunovic, Luca Beurer-Kellner, Marc Fischer, and Martin Vechev
    In ICML 2024 Next Generation of AI Safety Workshop, 2024
    PDF
  2. arXiv
    complai.png
    COMPL-AI Framework: A Technical Interpretation and LLM Benchmarking Suite for the EU Artificial Intelligence Act
    Philipp Guldimann, Alexander Spiridonov, Robin Staab, Nikola Jovanovic, Mark Vero, Velko Vechev, Anna Gueorguieva, Mislav Balunovic, Nikola Konstantinov, Pavol Bielik, and 2 more authors
    arXiv, 2024
    arXiv
  3. NeurIPS
    agentdojo.png
    AgentDojo: A Dynamic Environment to Evaluate Attacks and Defenses for LLM Agents
    Edoardo Debenedetti, Jie Zhang, Mislav Balunovic, Luca Beurer-Kellner, Marc Fischer, and Florian Tramèr
    In NeurIPS, 2024
    arXiv
  4. ICML
    cuts.png
    CuTS: Customizable Tabular Synthetic Data Generation
    Mark Vero, Mislav Balunovic, and Martin T. Vechev
    In ICML, 2024
    arXiv
  5. ICLR
    beyond_memorization.png
    Beyond Memorization: Violating Privacy via Inference with Large Language Models
    Robin Staab, Mark Vero, Mislav Balunovic, and Martin T. Vechev
    In ICLR, 2024
    Spotlight presentation arXiv Code
  6. IEEE S&P
    datamin.png
    From Principle to Practice: Vertical Data Minimization for Machine Learning
    Robin Staab, Nikola Jovanovic, Mislav Balunovic, and Martin T. Vechev
    In IEEE S&P, 2024
    arXiv

2023

  1. ICML
    fare.png
    FARE: Provably Fair Representation Learning with Practical Certificates
    Nikola Jovanovic, Mislav Balunovic, Dimitar Iliev Dimitrov, and Martin T. Vechev
    In ICML, 2023
    arXiv
  2. ICML
    tableak.png
    TabLeak: Tabular Data Leakage in Federated Learning
    Mark Vero, Mislav Balunovic, Dimitar Iliev Dimitrov, and Martin T. Vechev
    In ICML, 2023
    arXiv

2022

  1. NeurIPS
    lamp.png
    LAMP: Extracting Text from Gradients with Language Model Priors
    Mislav Balunovic, Dimitar I. Dimitrov, Nikola Jovanovic, and Martin T. Vechev
    In NeurIPS, 2022
    arXiv
  2. TMLR
    fedavg.png
    Data Leakage in Federated Averaging
    Dimitar Iliev Dimitrov, Mislav Balunovic, Nikola Konstantinov, and Martin T. Vechev
    Transactions of Machine Learning Research, 2022
    PDF
  3. ICLR
    bayes_framework.png
    Bayesian Framework for Gradient Leakage
    Mislav Balunovic, Dimitar Iliev Dimitrov, Robin Staab, and Martin T. Vechev
    In ICLR, 2022
    arXiv
  4. ECCV
    lassi.png
    Latent Space Smoothing for Individually Fair Representations
    Momchil Peychev, Anian Ruoss, Mislav Balunovic, Maximilian Baader, and Martin T. Vechev
    In ECCV, 2022
    arXiv
  5. TMLR
    paradox.png
    On the Paradox of Certified Training
    Nikola Jovanovic, Mislav Balunovic, Maximilian Baader, and Martin T. Vechev
    Transactions of Machine Learning Research, 2022
    arXiv
  6. ICLR
    fnf.png
    Fair Normalizing Flows
    Mislav Balunovic, Anian Ruoss, and Martin T. Vechev
    In ICLR, 2022
    arXiv

2021

  1. ICCV
    pointclouds.png
    Robustness Certification for Point Cloud Models
    Tobias Lorenz, Anian Ruoss, Mislav Balunovic, Gagandeep Singh, and Martin T. Vechev
    In ICCV, 2021
    arXiv
  2. CAV
    sound.png
    Scalable Polyhedral Verification of Recurrent Neural Networks
    Wonryong Ryou, Jiayu Chen, Mislav Balunovic, Gagandeep Singh, Andrei Marian Dan, and Martin T. Vechev
    In CAV, 2021
    arXiv
  3. ICLR
    compositional.png
    Certify or Predict: Boosting Certified Robustness with Compositional Architectures
    Mark Niklas Müller, Mislav Balunovic, and Martin T. Vechev
    In ICLR, 2021
    PDF
  4. AAAI
    spatial.png
    Efficient Certification of Spatial Robustness
    Anian Ruoss, Maximilian Baader, Mislav Balunovic, and Martin T. Vechev
    In AAAI, 2021
    arXiv

2020

  1. NeurIPS
    lcifr.png
    Learning Certified Individually Fair Representations
    Anian Ruoss, Mislav Balunovic, Marc Fischer, and Martin T. Vechev
    In NeurIPS, 2020
    arXiv
  2. ICLR
    colt.png
    Adversarial Training and Provable Defenses: Bridging the Gap
    Mislav Balunovic and Martin T. Vechev
    In ICLR, 2020
    Oral presentation PDF

2019

  1. NeurIPS
    geometric.png
    Certifying Geometric Robustness of Neural Networks
    Mislav Balunovic, Maximilian Baader, Gagandeep Singh, Timon Gehr, and Martin T. Vechev
    In NeurIPS, 2019
    PDF
  2. CCS
    fuzzing.png
    Learning to Fuzz from Symbolic Execution with Application to Smart Contracts
    Jingxuan He, Mislav Balunovic, Nodar Ambroladze, Petar Tsankov, and Martin T. Vechev
    In CCS, 2019
    PDF
  3. ICML
    dl2.png
    DL2: Training and Querying Neural Networks with Logic
    Marc Fischer, Mislav Balunovic, Dana Drachsler-Cohen, Timon Gehr, Ce Zhang, and Martin T. Vechev
    In ICML, 2019
    PDF

2018

  1. NeurIPS
    fastsmt.png
    Learning to Solve SMT Formulas
    Mislav Balunovic, Pavol Bielik, and Martin T. Vechev
    In NeurIPS, 2018
    Oral presentation PDF