Seguir
Dimitar I. Dimitrov
Dimitar I. Dimitrov
Ph.D. Student at ETH Zurich
Email confirmado em inf.ethz.ch - Página inicial
Título
Citado por
Citado por
Ano
Bayesian framework for gradient leakage
M Balunović, DI Dimitrov, R Staab, M Vechev
arXiv preprint arXiv:2111.04706, 2021
442021
Lamp: Extracting text from gradients with language model priors
M Balunovic, D Dimitrov, N Jovanović, M Vechev
Advances in Neural Information Processing Systems 35, 7641-7654, 2022
33*2022
Fast and precise certification of transformers
G Bonaert, DI Dimitrov, M Baader, M Vechev
Proceedings of the 42nd ACM SIGPLAN International Conference on Programming …, 2021
302021
Data leakage in federated averaging
DI Dimitrov, M Balunovic, N Konstantinov, M Vechev
Transactions on Machine Learning Research, 2022
142022
Shared certificates for neural network verification
M Fischer, C Sprecher, DI Dimitrov, G Singh, M Vechev
International Conference on Computer Aided Verification, 127-148, 2022
9*2022
TabLeak: Tabular data leakage in federated learning
M Vero, M Balunović, DI Dimitrov, M Vechev
Proceedings of the 40th International Conference on Machine Learning 202 …, 2023
82023
Provably robust adversarial examples
DI Dimitrov, G Singh, T Gehr, M Vechev
arXiv preprint arXiv:2007.12133, 2020
82020
Fare: Provably fair representation learning with practical certificates
N Jovanović, M Balunovic, DI Dimitrov, M Vechev
International Conference on Machine Learning, 15401-15420, 2023
6*2023
Group and Attack: Auditing Differential Privacy
J Lokna, A Paradis, DI Dimitrov, M Vechev
Proceedings of the 2023 ACM SIGSAC Conference on Computer and Communications …, 2023
12023
SPEAR: Exact Gradient Inversion of Batches in Federated Learning
DI Dimitrov, M Baader, MN Müller, M Vechev
arXiv preprint arXiv:2403.03945, 2024
2024
Hiding in Plain Sight: Disguising Data Stealing Attacks in Federated Learning
K Garov, DI Dimitrov, N Jovanović, M Vechev
arXiv preprint arXiv:2306.03013, 2023
2023
Image Inpainting with Gaussian Processes
DI Dimitrov
O sistema não pode efectuar a operação agora. Tente mais tarde.
Artigos 1–12