Follow
Mark Kurtz
Mark Kurtz
Neural Magic
Verified email at neuralmagic.com
Title
Cited by
Cited by
Year
Inducing and exploiting activation sparsity for fast inference on deep neural networks
M Kurtz, J Kopinsky, R Gelashvili, A Matveev, J Carr, M Goin, W Leiserson, ...
International Conference on Machine Learning, 5533-5543, 2020
1712020
The optimal bert surgeon: Scalable and accurate second-order pruning for large language models
E Kurtic, D Campos, T Nguyen, E Frantar, M Kurtz, B Fineran, M Goin, ...
arXiv preprint arXiv:2203.07259, 2022
1142022
How well do sparse imagenet models transfer?
E Iofinova, A Peste, M Kurtz, D Alistarh
Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern …, 2022
452022
Sparse* bert: Sparse models are robust
D Campos, A Marques, T Nguyen, M Kurtz, C Zhai
arXiv preprint arXiv:2205.12452, 2022
72022
Enabling High-Sparsity Foundational Llama Models with Efficient Pretraining and Deployment
A Agarwalla, A Gupta, A Marques, S Pandit, M Goin, E Kurtic, K Leong, ...
arXiv preprint arXiv:2405.03594, 2024
52024
System and method of accelerating execution of a neural network
A Matveev, D Alistarh, J Kopinsky, R Gelashvili, M Kurtz, N Shavit
US Patent 11,195,095, 2021
42021
oBERTa: Improving Sparse Transfer Learning via improved initialization, distillation, and pruning regimes
D Campos, A Marques, M Kurtz, CX Zhai
arXiv preprint arXiv:2303.17612, 2023
32023
Sparse* BERT: sparse models generalize to new tasks and domains
D Campos, A Marques, T Nguyen, M Kurtz, CX Zhai
arXiv preprint arXiv:2205.12452, 2022
22022
System and method of training a neural network
M Kurtz, D Alistarh
US Patent App. 17/149,043, 2021
12021
" Give Me BF16 or Give Me Death"? Accuracy-Performance Trade-Offs in LLM Quantization
E Kurtic, A Marques, S Pandit, M Kurtz, D Alistarh
arXiv preprint arXiv:2411.02355, 2024
2024
System and method of accelerating execution of a neural network
A Matveev, D Alistarh, J Kopinsky, R Gelashvili, M Kurtz, N Shavit
US Patent 11,797,855, 2023
2023
The system can't perform the operation now. Try again later.
Articles 1–11