Прати
Gintare Karolina Dziugaite
Gintare Karolina Dziugaite
Google DeepMind
Верификована је имејл адреса на google.com - Почетна страница
Наслов
Навело
Навело
Година
Computing nonvacuous generalization bounds for deep (stochastic) neural networks with many more parameters than training data
GK Dziugaite, DM Roy
arXiv preprint arXiv:1703.11008, 2017
8792017
Training generative neural networks via maximum mean discrepancy optimization
GK Dziugaite, DM Roy, Z Ghahramani
arXiv preprint arXiv:1505.03906, 2015
6972015
A study of the effect of jpg compression on adversarial images
GK Dziugaite, Z Ghahramani, DM Roy
arXiv preprint arXiv:1608.00853, 2016
6322016
Linear mode connectivity and the lottery ticket hypothesis
J Frankle, GK Dziugaite, D Roy, M Carbin
International Conference on Machine Learning, 3259-3269, 2020
6042020
Deep learning on a data diet: Finding important examples early in training
M Paul, S Ganguli, GK Dziugaite
Advances in neural information processing systems 34, 20596-20607, 2021
4582021
Stabilizing the lottery ticket hypothesis
J Frankle, GK Dziugaite, DM Roy, M Carbin
arXiv preprint arXiv:1903.01611, 2019
390*2019
Pruning neural networks at initialization: Why are we missing the mark?
J Frankle, GK Dziugaite, DM Roy, M Carbin
arXiv preprint arXiv:2009.08576, 2020
2662020
Deep learning versus kernel learning: an empirical study of loss landscape geometry and the time evolution of the neural tangent kernel
S Fort, GK Dziugaite, M Paul, S Kharaghani, DM Roy, S Ganguli
Advances in Neural Information Processing Systems 33, 5850-5861, 2020
2062020
Neural network matrix factorization
GK Dziugaite, DM Roy
arXiv preprint arXiv:1511.06443, 2015
1852015
Information-theoretic generalization bounds for SGLD via data-dependent estimates
J Negrea, M Haghifam, GK Dziugaite, A Khisti, DM Roy
Advances in Neural Information Processing Systems 32, 2019
1612019
Data-dependent PAC-Bayes priors via differential privacy
GK Dziugaite, DM Roy
Advances in neural information processing systems 31, 2018
1612018
Sharpened generalization bounds based on conditional mutual information and an application to noisy, iterative algorithms
M Haghifam, J Negrea, A Khisti, DM Roy, GK Dziugaite
Advances in Neural Information Processing Systems 33, 9925-9935, 2020
1192020
Relatif: Identifying explanatory training samples via relative influence
E Barshan, ME Brunet, GK Dziugaite
International Conference on Artificial Intelligence and Statistics, 1899-1909, 2020
1182020
On the role of data in PAC-Bayes bounds
GK Dziugaite, K Hsu, W Gharbieh, G Arpino, D Roy
International Conference on Artificial Intelligence and Statistics, 604-612, 2021
1172021
In search of robust measures of generalization
GK Dziugaite, A Drouin, B Neal, N Rajkumar, E Caballero, L Wang, ...
Advances in Neural Information Processing Systems 33, 11723-11733, 2020
1152020
Information-theoretic generalization bounds for stochastic gradient descent
G Neu, GK Dziugaite, M Haghifam, DM Roy
Conference on Learning Theory, 3526-3545, 2021
902021
Entropy-SGD optimizes the prior of a PAC-Bayes bound: Data-dependent PAC-Bayes priors via differential privacy
GK Dziugaite, DM Roy
74*2018
In defense of uniform convergence: Generalization via derandomization with an application to interpolating predictors
J Negrea, GK Dziugaite, D Roy
International Conference on Machine Learning, 7263-7272, 2020
692020
Enforcing interpretability and its statistical impacts: Trade-offs between accuracy and interpretability
GK Dziugaite, S Ben-David, DM Roy
arXiv preprint arXiv:2010.13764, 2020
642020
Neurips 2020 competition: Predicting generalization in deep learning
Y Jiang, P Foret, S Yak, DM Roy, H Mobahi, GK Dziugaite, S Bengio, ...
arXiv preprint arXiv:2012.07976, 2020
592020
Систем тренутно не може да изврши ову радњу. Пробајте поново касније.
Чланци 1–20