| A closer look at memorization in deep networks D Arpit, S Jastrzębski, N Ballas, D Krueger, E Bengio, MS Kanwal, ... ICML 2017 (arXiv preprint arXiv:1706.05394), 2017 | 2023 | 2017 |
| On the spectral bias of deep neural networks N Rahaman, D Arpit, A Baratin, F Draxler, M Lin, FA Hamprecht, Y Bengio, ... ICML 2019 (arXiv preprint arXiv:1806.08734), 2018 | 1365* | 2018 |
| Three factors influencing minima in SGD S Jastrzębski, Z Kenton, D Arpit, N Ballas, A Fischer, Y Bengio, A Storkey ICANN 2018 (arXiv preprint arXiv:1711.04623), 2017 | 529 | 2017 |
| The Break-Even Point on Optimization Trajectories of Deep Neural Networks S Jastrzebski, M Szymczak, S Fort, D Arpit, J Tabor, K Cho, K Geras ICLR 2020 (arXiv preprint arXiv:2002.09572), 2020 | 164 | 2020 |
| Residual connections encourage iterative inference S Jastrzebski, D Arpit, N Ballas, V Verma, T Che, Y Bengio ICLR 2018 (arXiv preprint arXiv:1710.04773), 2017 | 148 | 2017 |
| Normalization propagation: A parametric technique for removing internal covariate shift in deep networks D Arpit, Y Zhou, BU Kota, V Govindaraju ICML 2016 (arXiv preprint arXiv:1603.01431), 2016 | 146 | 2016 |
| Ensemble of averages: Improving model selection and boosting performance in domain generalization D Arpit, H Wang, Y Zhou, C Xiong NeurIPS 2022, 2021 | 126 | 2021 |
| A walk with sgd C Xing, D Arpit, C Tsirigotis, Y Bengio arXiv preprint arXiv:1802.08770, 2018 | 115 | 2018 |
| Why regularized auto-encoders learn sparse representation? D Arpit, Y Zhou, H Ngo, V Govindaraju ICML 2016 (arXiv preprint arXiv:1505.05561), 2015 | 93 | 2015 |
| Deep Nets Don't Learn via Memorization D Krueger, N Ballas, S Jastrzebski, D Arpit, MS Kanwal, T Maharaj, ... ICLR 2017 Workshop, 2017 | 73 | 2017 |
| How to Initialize your Network? Robust Initialization for WeightNorm & ResNets D Arpit, V Campos, Y Bengio NeurIPs 2019, 2019 | 64 | 2019 |
| Catastrophic Fisher Explosion: Early Phase Fisher Matrix Impacts Generalization S Jastrzebski, D Arpit, O Astrand, G Kerg, H Wang, C Xiong, R Socher, ... ICML 2021, 2020 | 62 | 2020 |
| Fraternal Dropout K Zolna, D Arpit, D Suhubdy, Y Bengio ICLR 2018 (arXiv preprint arXiv:1711.00066), 2017 | 62 | 2017 |
| Bolaa: Benchmarking and orchestrating llm-augmented autonomous agents Z Liu, W Yao, J Zhang, L Xue, S Heinecke, R Murthy, Y Feng, Z Chen, ... arXiv preprint arXiv:2308.05960, 2023 | 57 | 2023 |
| h-detach: Modifying the LSTM Gradient Towards Better Optimization D Arpit, B Kanuparthi, G Kerg, NR Ke, I Mitliagkas, Y Bengio ICLR 2019 (arXiv preprint arXiv:1810.03023), 2018 | 50 | 2018 |
| Variational bi-lstms S Shabanian, D Arpit, A Trischler, Y Bengio arXiv preprint arXiv:1711.05717, 2017 | 42 | 2017 |
| Retroformer: Retrospective large language agents with policy gradient optimization W Yao, S Heinecke, JC Niebles, Z Liu, Y Feng, L Xue, R Murthy, Z Chen, ... arXiv preprint arXiv:2308.02151, 2023 | 40 | 2023 |
| Is joint training better for deep auto-encoders? Y Zhou, D Arpit, I Nwogu, V Govindaraju arXiv preprint arXiv:1405.1380, 2014 | 39 | 2014 |
| Finding Flatter Minima with SGD S Jastrzębski, Z Kenton, D Arpit, N Ballas, A Fischer, Y Bengio, A Storkey ICLR 2018 Workshop, 2018 | 38 | 2018 |
| The benefits of over-parameterization at initialization in deep ReLU networks D Arpit, Y Bengio arXiv preprint arXiv:1901.03611, 2019 | 34 | 2019 |
Yoshua BengioProfessor of computer science, University of Montreal, Mila, IVADO, CIFARVerified email at umontreal.ca
