Giancarlo Kerg

13 papers A* 4Journal 8Unranked 1
YearRankTypeTitle / Venue / Authors
2024 J jnl
Neural networks with optimized single-neuron adaptation uncover biologically plausible regularization.
PLoS Comput. Biol.
Victor Geadah, Stefan Horoi, Giancarlo Kerg, Guy Wolf, Guillaume Lajoie
2022 A* conf
Continuous-Time Meta-Learning with Forward Mode Differentiation.
ICLR
Tristan Deleu, David Kanaa, Leo Feng, Giancarlo Kerg, Yoshua Bengio, Guillaume Lajoie, Pierre-Luc Bacon
2022 J jnl
Continuous-Time Meta-Learning with Forward Mode Differentiation.
CoRR
Tristan Deleu, David Kanaa, Leo Feng, Giancarlo Kerg, Yoshua Bengio, Guillaume Lajoie, Pierre-Luc Bacon
2022 J jnl
On Neural Architecture Inductive Biases for Relational Tasks.
CoRR
Giancarlo Kerg, Sarthak Mittal, David Rolnick, Yoshua Bengio, Blake A. Richards, Guillaume Lajoie
2021 A* conf
Catastrophic Fisher Explosion: Early Phase Fisher Matrix Impacts Generalization.
ICML
Stanislaw Jastrzebski, Devansh Arpit, Oliver Åstrand, Giancarlo Kerg, Huan Wang, Caiming Xiong, Richard Socher, Kyunghyun Cho, Krzysztof J. Geras
2020 J jnl
Advantages of biologically-inspired adaptive neural activation in RNNs during learning.
CoRR
Victor Geadah, Giancarlo Kerg, Stefan Horoi, Guy Wolf, Guillaume Lajoie
2020 J jnl
Catastrophic Fisher Explosion: Early Phase Fisher Matrix Impacts Generalization.
CoRR
Stanislaw Jastrzebski, Devansh Arpit, Oliver Åstrand, Giancarlo Kerg, Huan Wang, Caiming Xiong, Richard Socher, Kyunghyun Cho, Krzysztof J. Geras
2020 A* conf
Untangling tradeoffs between recurrence and self-attention in artificial neural networks.
NeurIPS
Giancarlo Kerg, Bhargav Kanuparthi, Anirudh Goyal, Kyle Goyette, Yoshua Bengio, Guillaume Lajoie
2020 J jnl
Untangling tradeoffs between recurrence and self-attention in neural networks.
CoRR
Giancarlo Kerg, Bhargav Kanuparthi, Anirudh Goyal, Kyle Goyette, Yoshua Bengio, Guillaume Lajoie
2019 A* conf
Non-normal Recurrent Neural Network (nnRNN): learning long time dependencies while improving expressivity with transient dynamics.
NeurIPS
Giancarlo Kerg, Kyle Goyette, Maximilian Puelma Touzel, Gauthier Gidel, Eugene Vorontsov, Yoshua Bengio, Guillaume Lajoie
2019 J jnl
Non-normal Recurrent Neural Network (nnRNN): learning long time dependencies while improving expressivity with transient dynamics.
CoRR
Giancarlo Kerg, Kyle Goyette, Maximilian Puelma Touzel, Gauthier Gidel, Eugene Vorontsov, Yoshua Bengio, Guillaume Lajoie
2019 — conf
h-detach: Modifying the LSTM Gradient Towards Better Optimization.
ICLR (Poster)
Bhargav Kanuparthi, Devansh Arpit, Giancarlo Kerg, Nan Rosemary Ke, Ioannis Mitliagkas, Yoshua Bengio
2018 J jnl
h-detach: Modifying the LSTM Gradient Towards Better Optimization.
CoRR
Devansh Arpit, Bhargav Kanuparthi, Giancarlo Kerg, Nan Rosemary Ke, Ioannis Mitliagkas, Yoshua Bengio