Alert button
Picture for Mohammad Pezeshki

Mohammad Pezeshki

Alert button

Feedback-guided Data Synthesis for Imbalanced Classification

Sep 29, 2023
Reyhane Askari Hemmat, Mohammad Pezeshki, Florian Bordes, Michal Drozdzal, Adriana Romero-Soriano

Viaarxiv icon

Discovering environments with XRM

Sep 28, 2023
Mohammad Pezeshki, Diane Bouchacourt, Mark Ibrahim, Nicolas Ballas, Pascal Vincent, David Lopez-Paz

Viaarxiv icon

Predicting Grokking Long Before it Happens: A look into the loss landscape of models which grok

Jun 23, 2023
Pascal Jr. Tikeng Notsawo, Hattie Zhou, Mohammad Pezeshki, Irina Rish, Guillaume Dumas

Figure 1 for Predicting Grokking Long Before it Happens: A look into the loss landscape of models which grok
Figure 2 for Predicting Grokking Long Before it Happens: A look into the loss landscape of models which grok
Figure 3 for Predicting Grokking Long Before it Happens: A look into the loss landscape of models which grok
Figure 4 for Predicting Grokking Long Before it Happens: A look into the loss landscape of models which grok
Viaarxiv icon

Multi-scale Feature Learning Dynamics: Insights for Double Descent

Dec 06, 2021
Mohammad Pezeshki, Amartya Mitra, Yoshua Bengio, Guillaume Lajoie

Figure 1 for Multi-scale Feature Learning Dynamics: Insights for Double Descent
Figure 2 for Multi-scale Feature Learning Dynamics: Insights for Double Descent
Figure 3 for Multi-scale Feature Learning Dynamics: Insights for Double Descent
Figure 4 for Multi-scale Feature Learning Dynamics: Insights for Double Descent
Viaarxiv icon

Simple data balancing achieves competitive worst-group-accuracy

Oct 27, 2021
Badr Youbi Idrissi, Martin Arjovsky, Mohammad Pezeshki, David Lopez-Paz

Figure 1 for Simple data balancing achieves competitive worst-group-accuracy
Figure 2 for Simple data balancing achieves competitive worst-group-accuracy
Figure 3 for Simple data balancing achieves competitive worst-group-accuracy
Figure 4 for Simple data balancing achieves competitive worst-group-accuracy
Viaarxiv icon

Gradient Starvation: A Learning Proclivity in Neural Networks

Nov 23, 2020
Mohammad Pezeshki, Sékou-Oumar Kaba, Yoshua Bengio, Aaron Courville, Doina Precup, Guillaume Lajoie

Figure 1 for Gradient Starvation: A Learning Proclivity in Neural Networks
Figure 2 for Gradient Starvation: A Learning Proclivity in Neural Networks
Figure 3 for Gradient Starvation: A Learning Proclivity in Neural Networks
Figure 4 for Gradient Starvation: A Learning Proclivity in Neural Networks
Viaarxiv icon

On the Learning Dynamics of Deep Neural Networks

Sep 18, 2018
Remi Tachet des Combes, Mohammad Pezeshki, Samira Shabanian, Aaron Courville, Yoshua Bengio

Figure 1 for On the Learning Dynamics of Deep Neural Networks
Figure 2 for On the Learning Dynamics of Deep Neural Networks
Figure 3 for On the Learning Dynamics of Deep Neural Networks
Figure 4 for On the Learning Dynamics of Deep Neural Networks
Viaarxiv icon

Negative Momentum for Improved Game Dynamics

Jul 12, 2018
Gauthier Gidel, Reyhane Askari Hemmat, Mohammad Pezeshki, Gabriel Huang, Remi Lepriol, Simon Lacoste-Julien, Ioannis Mitliagkas

Figure 1 for Negative Momentum for Improved Game Dynamics
Figure 2 for Negative Momentum for Improved Game Dynamics
Figure 3 for Negative Momentum for Improved Game Dynamics
Figure 4 for Negative Momentum for Improved Game Dynamics
Viaarxiv icon

Zoneout: Regularizing RNNs by Randomly Preserving Hidden Activations

Sep 22, 2017
David Krueger, Tegan Maharaj, János Kramár, Mohammad Pezeshki, Nicolas Ballas, Nan Rosemary Ke, Anirudh Goyal, Yoshua Bengio, Aaron Courville, Chris Pal

Figure 1 for Zoneout: Regularizing RNNs by Randomly Preserving Hidden Activations
Figure 2 for Zoneout: Regularizing RNNs by Randomly Preserving Hidden Activations
Figure 3 for Zoneout: Regularizing RNNs by Randomly Preserving Hidden Activations
Figure 4 for Zoneout: Regularizing RNNs by Randomly Preserving Hidden Activations
Viaarxiv icon

Towards End-to-End Speech Recognition with Deep Convolutional Neural Networks

Jan 10, 2017
Ying Zhang, Mohammad Pezeshki, Philemon Brakel, Saizheng Zhang, Cesar Laurent Yoshua Bengio, Aaron Courville

Figure 1 for Towards End-to-End Speech Recognition with Deep Convolutional Neural Networks
Figure 2 for Towards End-to-End Speech Recognition with Deep Convolutional Neural Networks
Figure 3 for Towards End-to-End Speech Recognition with Deep Convolutional Neural Networks
Figure 4 for Towards End-to-End Speech Recognition with Deep Convolutional Neural Networks
Viaarxiv icon