Les inscriptions sont ouvertes pour notre programme de 4 mois à temps plein, à Mila, conçu spécifiquement pour les fondateurs et fondatrices de la deep tech issus de parcours d'élite en STIM. Vous avez jusqu'au 21 novembre 2026 pour appliquer!
La prochaine rencontre, qui aura lieu le 10 novembre à Mila, permettra d'explorer comment pouvons-nous collectivement développer, encadrer et déployer des systèmes agentiques performants, fiables et sécuritaires en connectant chercheur·euse·s académiques, expert·e·s industriel·le·s et praticien·ne·s.
Le processus de demandes de supervision à Mila est maintenant ouvert! Vous pouvez appliquer dès maintenant à la maitrise de recherche ou Doctorat et joindre notre grande communauté de chercheur·euse·s à Mila pour l’automne 2027. Date limite : 1er décembre 2026
Nous utilisons des témoins pour analyser le trafic et l’utilisation de notre site web, afin de personnaliser votre expérience. Vous pouvez désactiver ces technologies à tout moment, mais cela peut restreindre certaines fonctionnalités du site. Consultez notre Politique de protection de la vie privée pour en savoir plus.
Paramètre des cookies
Vous pouvez activer et désactiver les types de cookies que vous souhaitez accepter. Cependant certains choix que vous ferez pourraient affecter les services proposés sur nos sites (ex : suggestions, annonces personnalisées, etc.).
Cookies essentiels
Ces cookies sont nécessaires au fonctionnement du site et ne peuvent être désactivés. (Toujours actif)
Cookies analyse
Acceptez-vous l'utilisation de cookies pour mesurer l'audience de nos sites ?
Lecteur Multimédia
Acceptez-vous l'utilisation de cookies pour afficher et vous permettre de regarder les contenus vidéo hébergés par nos partenaires (YouTube, etc.) ?
Humans are able to accelerate their learning by selecting training materials that are the most informative and at the appropriate level of d… (voir plus)ifficulty. We propose a framework for distributing deep learning in which one set of workers search for the most informative examples in parallel while a single worker updates the model on examples selected by importance sampling. This leads the model to update using an unbiased estimate of the gradient which also has minimum variance when the sampling proposal is proportional to the L2-norm of the gradient. We show experimentally that this method reduces gradient variance even in a context where the cost of synchronization across machines cannot be ignored, and where the factors for importance sampling are not updated instantly across the training set.
Associative memories allow the retrieval of previously stored messages given a part of their content. In this paper, we are interested in as… (voir plus)sociative memories based on c-partite graphs that were recently introduced. These memories are almost optimal in terms of the amount of storage they require (efficiency) and allow retrieving messages with low complexity. We propose a generic implementation model for the retrieval algorithm that can be readily mapped to an integrated circuit and study the retrieval performance when hardware components are affected by faults. We show using analytical and simulation results that these associative memories can be made resilient to circuit faults with a minor modification of the retrieval algorithm. In one example, the memory retains 88% of its efficiency when 1% of the storage cells are faulty, or 98% when 0.1% of the binary outputs of the retrieval algorithm are faulty. When considering storage faults, the fault tolerance exhibited by the proposed associative memory can be comparable to using a capacity-achieving error correction code for protecting the stored information.
We explore the question of whether the representations learned by classifiers can be used to enhance the quality of generative models. Our c… (voir plus)onjecture is that labels correspond to characteristics of natural data which are most salient to humans: identity in faces, objects in images, and utterances in speech. We propose to take advantage of this by using the representations from discriminative classifiers to augment the objective function corresponding to a generative model. In particular we enhance the objective function of the variational autoencoder, a popular generative model, with a discriminative regularization term. We show that enhancing the objective function in this way leads to samples that are clearer and have higher visual quality than the samples from the standard variational autoencoders.
. While deep neural nets (DNN’s) achieve impressive performance on image recognition tasks, previous studies have reported that DNN’s gi… (voir plus)ve high confidence predictions for unrecognizable images. Motivated by the observation that such fooling examples might be caused by the extrapolating nature of the log-softmax, we propose to combine neural networks with Learning Vector Quantization (LVQ). Our proposed method, called Deep LVQ (DLVQ), achieves comparable performance on MNIST while being more robust against fooling and adversarial examples.
2015-12-31
The European Symposium on Artificial Neural Networks (publié)
Many of the current state-of-the-art Large Vocabulary Continuous Speech Recognition Systems (LVCSR) are hybrids of neural networks and Hidde… (voir plus)n Markov Models (HMMs). Most of these systems contain separate components that deal with the acoustic modelling, language modelling and sequence decoding. We investigate a more direct approach in which the HMM is replaced with a Recurrent Neural Network (RNN) that performs sequence prediction directly at the character level. Alignment between the input features and the desired character sequence is learned automatically by an attention mechanism built into the RNN. For each predicted character, the attention mechanism scans the input sequence and chooses relevant frames. We propose two methods to speed up this operation: limiting the scan to a subset of most promising frames and pooling over time the information contained in neighboring frames, thereby reducing source sequence length. Integrating an n-gram language model into the decoding process yields recognition accuracies similar to other HMM-free RNN-based approaches.
2015-12-31
IEEE International Conference on Acoustics, Speech and Signal Processing (publié)
Medium, Moore Data, Carly Strasser from June 07, 2016 Open access to research articles has been in the news quite a bit lately (see the SciH… (voir plus)ub controversy, the preprints in biology discussion, and the European Union’s recent announcement). The Data-Driven Discovery team at the Moore Foundation has also been discussing open access, particularly as it relates to the publications generated by our #MooreData researchers. Our grantee population is fairly progressive when it comes to open science, and many of the outputs that they generate are already publicly available (including proposals, software, workflows, and publications). It is therefore easy for us to imagine that they would embrace a policy that mandates open access for research articles that they produce. That said, we are always open to discussions!
Memory networks are neural networks with an explicit memory component that can be both read and written to by the network. The memory is oft… (voir plus)en addressed in a soft way using a softmax function, making end-to-end training with backpropagation possible. However, this is not computationally scalable for applications which require the network to read from extremely large memories. On the other hand, it is well known that hard attention mechanisms based on reinforcement learning are challenging to train successfully. In this paper, we explore a form of hierarchical memory network, which can be considered as a hybrid between hard and soft attention memory networks. The memory is organized in a hierarchical structure such that reading from it is done with less computation than soft attention over a flat memory, while also being easier to train than hard attention over a flat memory. Specifically, we propose to incorporate Maximum Inner Product Search (MIPS) in the training and inference procedures for our hierarchical memory network. We explore the use of various state-of-the art approximate MIPS techniques and report results on SimpleQuestions, a challenging large scale factoid question answering task.
The Teacher Forcing algorithm trains recurrent networks by supplying observed sequence values as inputs during training and using the networ… (voir plus)k’s own one-step-ahead predictions to do multi-step sampling. We introduce the Professor Forcing algorithm, which uses adversarial domain adaptation to encourage the dynamics of the recurrent network to be the same when training the network and when sampling from the network over multiple time steps. We apply Professor Forcing to language modeling, vocal synthesis on raw waveforms, handwriting generation, and image generation. Empirically we find that Professor Forcing acts as a regularizer, improving test likelihood on character level Penn Treebank and sequential MNIST. We also find that the model qualitatively improves samples, especially when sampling for a large number of time steps. This is supported by human evaluation of sample quality. Trade-offs between Professor Forcing and Scheduled Sampling are discussed. We produce T-SNEs showing that Professor Forcing successfully makes the dynamics of the network during training and sampling more similar.
2015-12-31
Advances in Neural Information Processing Systems 29 (NIPS 2016) (publié)
Theano is a Python library that allows to define, optimize, and evaluate mathematical expressions involving multi-dimensional arrays efficie… (voir plus)ntly. Since its introduction, it has been one of the most used CPU and GPU mathematical compilers - especially in the machine learning community - and has shown steady performance improvements. Theano is being actively and continuously developed since 2008, multiple frameworks have been built on top of it and it has been used to produce many state-of-the-art machine learning models. The present article is structured as follows. Section I provides an overview of the Theano software and its community. Section II presents the principal features of Theano and how to use them, and compares them with other similar projects. Section III focuses on recently-introduced functionalities and improvements. Section IV compares the performance of Theano against Torch7 and TensorFlow on several machine learning models. Section V discusses current limitations of Theano and potential ways of improving it.