作者
Yoshua Bengio
Machine Learning
Compositional generalization in a deep seq2seq model by separating syntax and semantics
Comments 18 pages, 15 figures, preprint version of submission to NeurIPS 2019, under review
The Journey is the Reward: Unsupervised Learning of Influential Trajectories
Comments ICML'19 ERL Workshop
Low-memory convolutional neural networks through incremental depth-first processing
Manifold Mixup: Better Representations by Interpolating Hidden States
Comments To appear in ICML 2019
Visualizing the Consequences of Climate Change Using Cycle-Consistent Adversarial Networks
Learning Problem-agnostic Speech Representations from Multiple Self-supervised Tasks
Learning Speaker Representations with Mutual Information
Comments Submitted to Interspeech 2019
Towards Standardization of Data Licenses: The Montreal Data License
Wasserstein Dependency Measure for Representation Learning
Learning Dynamics Model in Reinforcement Learning by Incorporating the Long Term Future
Comments To appear at ICLR 2019
Generalization in Machine Learning via Analytical Learning Theory
Hyperbolic Discounting and Learning over Multiple Horizons
Speech and Speaker Recognition from Raw Waveform with SincNet
Comments arXiv admin note: substantial text overlap with arXiv:1811.09725, arXiv:1808.00158
Learning deep representations by mutual information estimation and maximization
Comments Accepted as an oral presentation at the International Conference for Learning Representations (ICLR), 2019
Towards Non-saturating Recurrent Units for Modelling Long-term Dependencies
Comments In Proceedings of AAAI 2019
The PyTorch-Kaldi Speech Recognition Toolkit
Comments Accepted at ICASSP 2019
A Meta-Transfer Objective for Learning to Disentangle Causal Mechanisms
Recall Traces: Backtracking Models for Efficient Reinforcement Learning
Comments Accepted at ICLR 2019
Adversarial Domain Adaptation for Stable Brain-Machine Interfaces
Comments 14 pages, 6 figures
h-detach: Modifying the LSTM Gradient Towards Better Optimization
Comments First two authors contributed equally. Published in ICLR 2019
Quaternion Recurrent Neural Networks
Comments ICLR Update - Full rework
Quantized Guided Pruning for Efficient Hardware Implementations of Convolutional Neural Networks
Deep Graph Infomax
Comments To appear at ICLR 2019. 17 pages, 8 figures
Towards the Latent Transcriptome
Comments 7 figures
The effects of negative adaptation in Model-Agnostic Meta-Learning
Comments Workshop on Meta-Learning - 32nd Conference on Neural Information Processing Systems (NeurIPS 2018), Montreal, Canada
DEFactor: Differentiable Edge Factorization-based Probabilistic Graph Generation
Representation Mixing for TTS Synthesis
Comments 5 pages, 3 figures
Deep convolutional networks for quality assessment of protein folds
Comments 8 pages
