Get our free extension to see links to code for papers anywhere online!

Add to Chrome

Add to Firefox

Picture for Martin Jaggi

Martin Jaggi

EPFL

Understanding Memorization from the Perspective of Optimization via Efficient Influence Estimation

Dec 16, 2021
Futong Liu, Tao Lin, Martin Jaggi

Figure 1 for Understanding Memorization from the Perspective of Optimization via Efficient Influence Estimation

Figure 2 for Understanding Memorization from the Perspective of Optimization via Efficient Influence Estimation

Figure 3 for Understanding Memorization from the Perspective of Optimization via Efficient Influence Estimation

Figure 4 for Understanding Memorization from the Perspective of Optimization via Efficient Influence Estimation

Access Paper or Ask Questions

Interpreting Language Models Through Knowledge Graph Extraction

Nov 16, 2021
Vinitra Swamy, Angelika Romanou, Martin Jaggi

Figure 1 for Interpreting Language Models Through Knowledge Graph Extraction

Figure 2 for Interpreting Language Models Through Knowledge Graph Extraction

Figure 3 for Interpreting Language Models Through Knowledge Graph Extraction

Figure 4 for Interpreting Language Models Through Knowledge Graph Extraction

* Published at NeurIPS 2021: eXplainable AI for Debugging and Diagnosis Workshop

Access Paper or Ask Questions

Linear Speedup in Personalized Collaborative Learning

Nov 10, 2021
El Mahdi Chayti, Sai Praneeth Karimireddy, Sebastian U. Stich, Nicolas Flammarion, Martin Jaggi

Figure 1 for Linear Speedup in Personalized Collaborative Learning

Figure 2 for Linear Speedup in Personalized Collaborative Learning

Figure 3 for Linear Speedup in Personalized Collaborative Learning

Figure 4 for Linear Speedup in Personalized Collaborative Learning

Access Paper or Ask Questions

Optimal Model Averaging: Towards Personalized Collaborative Learning

Oct 25, 2021
Felix Grimberg, Mary-Anne Hartley, Sai P. Karimireddy, Martin Jaggi

Figure 1 for Optimal Model Averaging: Towards Personalized Collaborative Learning

Figure 2 for Optimal Model Averaging: Towards Personalized Collaborative Learning

Figure 3 for Optimal Model Averaging: Towards Personalized Collaborative Learning

* 9 pages (12 pages incl. references and appendix), 1 figure, Best Paper at International Workshop on Federated Learning for User Privacy and Data Confidentiality in Conjunction with ICML 2021 (FL-ICML'21) ( https://web.archive.org/web/20210908135923/http://federated-learning.org/fl-icml-2021/ICML\%202021\%20Best\%20Paper.pdf )

Access Paper or Ask Questions

WAFFLE: Weighted Averaging for Personalized Federated Learning

Oct 13, 2021
Martin Beaussart, Felix Grimberg, Mary-Anne Hartley, Martin Jaggi

Figure 1 for WAFFLE: Weighted Averaging for Personalized Federated Learning

Figure 2 for WAFFLE: Weighted Averaging for Personalized Federated Learning

Figure 3 for WAFFLE: Weighted Averaging for Personalized Federated Learning

Figure 4 for WAFFLE: Weighted Averaging for Personalized Federated Learning

* Submitted to NeurIPS 2021 Workshop on New Frontiers in Federated Learning: Privacy, Fairness, Robustness, Personalization and Data Ownership

Access Paper or Ask Questions

RelaySum for Decentralized Deep Learning on Heterogeneous Data

Oct 08, 2021
Thijs Vogels, Lie He, Anastasia Koloskova, Tao Lin, Sai Praneeth Karimireddy, Sebastian U. Stich, Martin Jaggi

Figure 1 for RelaySum for Decentralized Deep Learning on Heterogeneous Data

Figure 2 for RelaySum for Decentralized Deep Learning on Heterogeneous Data

Figure 3 for RelaySum for Decentralized Deep Learning on Heterogeneous Data

Figure 4 for RelaySum for Decentralized Deep Learning on Heterogeneous Data

* To appear in NeurIPS 2021

Access Paper or Ask Questions

On Second-order Optimization Methods for Federated Learning

Sep 06, 2021
Sebastian Bischoff, Stephan Günnemann, Martin Jaggi, Sebastian U. Stich

Figure 1 for On Second-order Optimization Methods for Federated Learning

Figure 2 for On Second-order Optimization Methods for Federated Learning

Figure 3 for On Second-order Optimization Methods for Federated Learning

* ICML 2021 Workshop "Beyond first-order methods in ML systems"

Access Paper or Ask Questions

Semantic Perturbations with Normalizing Flows for Improved Generalization

Aug 18, 2021
Oguz Kaan Yuksel, Sebastian U. Stich, Martin Jaggi, Tatjana Chavdarova

Figure 1 for Semantic Perturbations with Normalizing Flows for Improved Generalization

Figure 2 for Semantic Perturbations with Normalizing Flows for Improved Generalization

Figure 3 for Semantic Perturbations with Normalizing Flows for Improved Generalization

Figure 4 for Semantic Perturbations with Normalizing Flows for Improved Generalization

* In Proceedings of the IEEE International Conference on Computer Vision

Access Paper or Ask Questions

A Field Guide to Federated Optimization

Jul 14, 2021
Jianyu Wang, Zachary Charles, Zheng Xu, Gauri Joshi, H. Brendan McMahan, Blaise Aguera y Arcas, Maruan Al-Shedivat, Galen Andrew, Salman Avestimehr, Katharine Daly, Deepesh Data, Suhas Diggavi, Hubert Eichner, Advait Gadhikar, Zachary Garrett, Antonious M. Girgis, Filip Hanzely, Andrew Hard, Chaoyang He, Samuel Horvath, Zhouyuan Huo, Alex Ingerman, Martin Jaggi, Tara Javidi, Peter Kairouz, Satyen Kale, Sai Praneeth Karimireddy, Jakub Konecny, Sanmi Koyejo, Tian Li, Luyang Liu, Mehryar Mohri, Hang Qi, Sashank J. Reddi, Peter Richtarik, Karan Singhal, Virginia Smith, Mahdi Soltanolkotabi, Weikang Song, Ananda Theertha Suresh, Sebastian U. Stich, Ameet Talwalkar, Hongyi Wang, Blake Woodworth, Shanshan Wu, Felix X. Yu, Honglin Yuan, Manzil Zaheer, Mi Zhang, Tong Zhang, Chunxiang Zheng, Chen Zhu, Wennan Zhu

Figure 1 for A Field Guide to Federated Optimization

Figure 2 for A Field Guide to Federated Optimization

Figure 3 for A Field Guide to Federated Optimization

Figure 4 for A Field Guide to Federated Optimization

Access Paper or Ask Questions

IFedAvg: Interpretable Data-Interoperability for Federated Learning

Jul 14, 2021
David Roschewitz, Mary-Anne Hartley, Luca Corinzia, Martin Jaggi

Figure 1 for IFedAvg: Interpretable Data-Interoperability for Federated Learning

Figure 2 for IFedAvg: Interpretable Data-Interoperability for Federated Learning

Figure 3 for IFedAvg: Interpretable Data-Interoperability for Federated Learning

Figure 4 for IFedAvg: Interpretable Data-Interoperability for Federated Learning

Access Paper or Ask Questions

Implicit Gradient Alignment in Distributed and Federated Learning

Jun 25, 2021
Yatin Dandi, Luis Barba, Martin Jaggi

Figure 1 for Implicit Gradient Alignment in Distributed and Federated Learning

Figure 2 for Implicit Gradient Alignment in Distributed and Federated Learning

Figure 3 for Implicit Gradient Alignment in Distributed and Federated Learning

Figure 4 for Implicit Gradient Alignment in Distributed and Federated Learning

Access Paper or Ask Questions

Simultaneous Training of Partially Masked Neural Networks

Jun 16, 2021
Amirkeivan Mohtashami, Martin Jaggi, Sebastian U. Stich

Figure 1 for Simultaneous Training of Partially Masked Neural Networks

Figure 2 for Simultaneous Training of Partially Masked Neural Networks

Figure 3 for Simultaneous Training of Partially Masked Neural Networks

Figure 4 for Simultaneous Training of Partially Masked Neural Networks

Access Paper or Ask Questions

Lightweight Cross-Lingual Sentence Representation Learning

Jun 12, 2021
Zhuoyuan Mao, Prakhar Gupta, Chenhui Chu, Martin Jaggi, Sadao Kurohashi

Figure 1 for Lightweight Cross-Lingual Sentence Representation Learning

Figure 2 for Lightweight Cross-Lingual Sentence Representation Learning

Figure 3 for Lightweight Cross-Lingual Sentence Representation Learning

Figure 4 for Lightweight Cross-Lingual Sentence Representation Learning

* ACL 2021

Access Paper or Ask Questions

Obtaining Better Static Word Embeddings Using Contextual Embedding Models

Jun 08, 2021
Prakhar Gupta, Martin Jaggi

Figure 1 for Obtaining Better Static Word Embeddings Using Contextual Embedding Models

Figure 2 for Obtaining Better Static Word Embeddings Using Contextual Embedding Models

Figure 3 for Obtaining Better Static Word Embeddings Using Contextual Embedding Models

Figure 4 for Obtaining Better Static Word Embeddings Using Contextual Embedding Models

* ACL 2021 accept

Access Paper or Ask Questions

Federated Learning for Malware Detection in IoT Devices

Apr 15, 2021
Valerian Rey, Pedro Miguel Sánchez Sánchez, Alberto Huertas Celdrán, Gérôme Bovet, Martin Jaggi

Figure 1 for Federated Learning for Malware Detection in IoT Devices

Figure 2 for Federated Learning for Malware Detection in IoT Devices

Figure 3 for Federated Learning for Malware Detection in IoT Devices

Figure 4 for Federated Learning for Malware Detection in IoT Devices

Access Paper or Ask Questions

Critical Parameters for Scalable Distributed Learning with Large Batches and Asynchronous Updates

Mar 03, 2021
Sebastian U. Stich, Amirkeivan Mohtashami, Martin Jaggi

Figure 1 for Critical Parameters for Scalable Distributed Learning with Large Batches and Asynchronous Updates

Figure 2 for Critical Parameters for Scalable Distributed Learning with Large Batches and Asynchronous Updates

Figure 3 for Critical Parameters for Scalable Distributed Learning with Large Batches and Asynchronous Updates

Figure 4 for Critical Parameters for Scalable Distributed Learning with Large Batches and Asynchronous Updates

* Proceedings of the 24th International Conference on Artificial Intelligence and Statistics (AISTATS) 2021

Access Paper or Ask Questions

Consensus Control for Decentralized Deep Learning

Feb 09, 2021
Lingjing Kong, Tao Lin, Anastasia Koloskova, Martin Jaggi, Sebastian U. Stich

Figure 1 for Consensus Control for Decentralized Deep Learning

Figure 2 for Consensus Control for Decentralized Deep Learning

Figure 3 for Consensus Control for Decentralized Deep Learning

Figure 4 for Consensus Control for Decentralized Deep Learning

* LK and TL contribute equally

Access Paper or Ask Questions

Quasi-Global Momentum: Accelerating Decentralized Deep Learning on Heterogeneous Data

Feb 09, 2021
Tao Lin, Sai Praneeth Karimireddy, Sebastian U. Stich, Martin Jaggi

Figure 1 for Quasi-Global Momentum: Accelerating Decentralized Deep Learning on Heterogeneous Data

Figure 2 for Quasi-Global Momentum: Accelerating Decentralized Deep Learning on Heterogeneous Data

Figure 3 for Quasi-Global Momentum: Accelerating Decentralized Deep Learning on Heterogeneous Data

Figure 4 for Quasi-Global Momentum: Accelerating Decentralized Deep Learning on Heterogeneous Data

Access Paper or Ask Questions

Exact Optimization of Conformal Predictors via Incremental and Decremental Learning

Feb 05, 2021
Giovanni Cherubin, Konstantinos Chatzikokolakis, Martin Jaggi

Figure 1 for Exact Optimization of Conformal Predictors via Incremental and Decremental Learning

Figure 2 for Exact Optimization of Conformal Predictors via Incremental and Decremental Learning

Figure 3 for Exact Optimization of Conformal Predictors via Incremental and Decremental Learning

Figure 4 for Exact Optimization of Conformal Predictors via Incremental and Decremental Learning

Access Paper or Ask Questions

Learning from History for Byzantine Robust Optimization

Dec 18, 2020
Sai Praneeth Karimireddy, Lie He, Martin Jaggi

Figure 1 for Learning from History for Byzantine Robust Optimization

Figure 2 for Learning from History for Byzantine Robust Optimization

Figure 3 for Learning from History for Byzantine Robust Optimization

Figure 4 for Learning from History for Byzantine Robust Optimization

* 22 pages, 5 figures

Access Paper or Ask Questions

A Linearly Convergent Algorithm for Decentralized Optimization: Sending Less Bits for Free!

Nov 03, 2020
Dmitry Kovalev, Anastasia Koloskova, Martin Jaggi, Peter Richtarik, Sebastian U. Stich

Figure 1 for A Linearly Convergent Algorithm for Decentralized Optimization: Sending Less Bits for Free!

Figure 2 for A Linearly Convergent Algorithm for Decentralized Optimization: Sending Less Bits for Free!

Figure 3 for A Linearly Convergent Algorithm for Decentralized Optimization: Sending Less Bits for Free!

Figure 4 for A Linearly Convergent Algorithm for Decentralized Optimization: Sending Less Bits for Free!

Access Paper or Ask Questions

Sparse Communication for Training Deep Networks

Sep 19, 2020
Negar Foroutan Eghlidi, Martin Jaggi

Figure 1 for Sparse Communication for Training Deep Networks

Figure 2 for Sparse Communication for Training Deep Networks

Figure 3 for Sparse Communication for Training Deep Networks

Access Paper or Ask Questions

PowerGossip: Practical Low-Rank Communication Compression in Decentralized Deep Learning

Aug 04, 2020
Thijs Vogels, Sai Praneeth Karimireddy, Martin Jaggi

Access Paper or Ask Questions

Multi-Head Attention: Collaborate Instead of Concatenate

Jun 29, 2020
Jean-Baptiste Cordonnier, Andreas Loukas, Martin Jaggi

Figure 1 for Multi-Head Attention: Collaborate Instead of Concatenate

Figure 2 for Multi-Head Attention: Collaborate Instead of Concatenate

Figure 3 for Multi-Head Attention: Collaborate Instead of Concatenate

Figure 4 for Multi-Head Attention: Collaborate Instead of Concatenate

Access Paper or Ask Questions

Taming GANs with Lookahead

Jun 25, 2020
Tatjana Chavdarova, Matteo Pagliardini, Martin Jaggi, Francois Fleuret

Figure 1 for Taming GANs with Lookahead

Figure 2 for Taming GANs with Lookahead

Figure 3 for Taming GANs with Lookahead

Figure 4 for Taming GANs with Lookahead

Access Paper or Ask Questions

Byzantine-Robust Learning on Heterogeneous Datasets via Resampling

Jun 23, 2020
Lie He, Sai Praneeth Karimireddy, Martin Jaggi

Figure 1 for Byzantine-Robust Learning on Heterogeneous Datasets via Resampling

Figure 2 for Byzantine-Robust Learning on Heterogeneous Datasets via Resampling

Figure 3 for Byzantine-Robust Learning on Heterogeneous Datasets via Resampling

Figure 4 for Byzantine-Robust Learning on Heterogeneous Datasets via Resampling

Access Paper or Ask Questions

Dynamic Model Pruning with Feedback

Jun 12, 2020
Tao Lin, Sebastian U. Stich, Luis Barba, Daniil Dmitriev, Martin Jaggi

Figure 1 for Dynamic Model Pruning with Feedback

Figure 2 for Dynamic Model Pruning with Feedback

Figure 3 for Dynamic Model Pruning with Feedback

Figure 4 for Dynamic Model Pruning with Feedback

* appearing at ICLR 2020

Access Paper or Ask Questions