Mohammad Pezeshki
Scholar

Mohammad Pezeshki

Google Scholar ID: HT85tXsAAAAJ
Mila
Robust Machine Learning
Citations & Impact
All-time
Citations
2,101
 
H-index
13
 
i10-index
15
 
Publications
20
 
Co-authors
19
list available
Resume (English only)
Academic Achievements
  • Multi-scale Feature Learning Dynamics: Insights for Double Descent – Analyzed epoch-wise double descent using statistical physics tools and derived analytical expressions for generalization error evolution
  • Gradient Starvation: A Learning Proclivity in Neural Networks – Identified and formalized 'Gradient Starvation', a phenomenon in over-parameterized neural networks
  • On the Learning Dynamics of Deep Neural Networks – Proved sigmoidal shape of classification error in deep networks under linear separability assumptions
  • Deconstructing the Ladder Network Architecture – Conducted ablation studies showing lateral connections are most critical in Ladder Networks
  • Negative momentum for improved game dynamics – Proposed alternating gradient updates with negative momentum for stable convergence in GANs
  • Zoneout: Regularizing RNNs by Randomly Preserving Hidden Activations – Introduced Zoneout, a stochastic regularization method for RNNs