Le traitement du langage naturel à l'ère de l'IA générative
Rejoignez-nous à Mila en octobre pour un atelier de trois jour visant à explorer le potentiel de transformation des technologies langagières et leurs implications pour la société.
Ce programme est conçu pour fournir aux professionnel·le·s travaillant dans le domaine de la politique une compréhension fondamentale de la technologie de l'IA.
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 ?
Multimedia Player
Acceptez-vous l'utilisation de cookies pour afficher et vous permettre de regarder les contenus vidéo hébergés par nos partenaires (YouTube, etc.) ?
Publications
Goal-driven optimization of single-neuron properties in artificial networks reveals regularization role of neural diversity and adaptation in the brain
Neurons in the brain have rich and adaptive input-output properties. Features such as diverse f-I curves and spike frequency adaptation are … (voir plus)known to place single neurons in optimal coding regimes when facing changing stimuli. Yet, it is still unclear how brain circuits exploit single neuron flexibility, and how network-level requirements may have shaped such cellular function. To answer this question, a multi-scaled approach is needed where the computations of single neurons and of neural circuits must be considered as a complete system. In this work, we use artificial neural networks to systematically investigate single neuron input-output adaptive mechanisms, optimized in an end-to-end fashion. Throughout the optimization process, each neuron has the liberty to modify its nonlinear activation function, parametrized to mimic f-I curves of biological neurons, and to learn adaptation strategies to modify activation functions in real-time during a task. We find that such networks show much-improved robustness to noise and changes in input statistics. Importantly, we find that this procedure recovers precise coding strategies found in biological neurons, such as gain scaling and fractional order differentiation/integration. Using tools from dynamical systems theory, we analyze the role of these emergent single neuron properties and argue that neural diversity and adaptation plays an active regularization role that enables neural circuits to optimally propagate information across time.
The study of first-order optimization is sensitive to the assumptions made on the objective functions.
These assumptions induce complexity c… (voir plus)lasses which play a key role in worst-case analysis, including
the fundamental concept of algorithm optimality. Recent work argues that strong convexity and
smoothness—popular assumptions in literature—lead to a pathological definition of the condition
number. Motivated by this result, we focus on the class of functions
satisfying a lower restricted secant inequality and an upper error bound. On top of being robust to
the aforementioned pathological behavior and including some non-convex functions, this pair of
conditions displays interesting geometrical properties. In particular, the necessary and sufficient
conditions to interpolate a set of points and their gradients within the class can be separated into
simple conditions on each sampled gradient. This allows the performance estimation problem (PEP)
to be solved analytically, leading to a lower bound
on the convergence rate that proves gradient descent to be exactly optimal on this class of functions
among all first-order algorithms.
Plants are dynamic systems that are integral to our existence and survival. Plants face environment changes and adapt over time to their sur… (voir plus)rounding conditions. We argue that plant responses to an environmental stimulus are a good example of a real-world problem that can be approached within a reinforcement learning (RL)framework. With the objective of controlling a plant by moving the light source, we propose GrowSpace, as a new RL benchmark. The back-end of the simulator is implemented using the Space Colonisation Algorithm, a plant growing model based on competition for space. Compared to video game RL environments, this simulator addresses a real-world problem and serves as a test bed to visualize plant growth and movement in a faster way than physical experiments. GrowSpace is composed of a suite of challenges that tackle several problems such as control, multi-stage learning,fairness and multi-objective learning. We provide agent baselines alongside case studies to demonstrate the difficulty of the proposed benchmark.
Reliable evaluation benchmarks designed for replicability and comprehensiveness have driven progress in machine learning. Due to the lack of… (voir plus) a multilingual benchmark, however, vision-and-language research has mostly focused on English language tasks. To fill this gap, we introduce the Image-Grounded Language Understanding Evaluation benchmark. IGLUE brings together - by both aggregating pre-existing datasets and creating new ones - visual question answering, cross-modal retrieval, grounded reasoning, and grounded entailment tasks across 20 diverse languages. Our benchmark enables the evaluation of multilingual multimodal models for transfer learning, not only in a zero-shot setting, but also in newly defined few-shot learning setups. Based on the evaluation of the available state-of-the-art models, we find that translate-test transfer is superior to zero-shot transfer and that few-shot learning is hard to harness for many tasks. Moreover, downstream performance is partially explained by the amount of available unlabelled textual data for pretraining, and only weakly by the typological distance of target-source languages. We hope to encourage future research efforts in this area by releasing the benchmark to the community.
Stochastic gradient descent (SGD) is a pillar of modern machine learning, serving as the go-to optimization algorithm for a diverse array of… (voir plus) problems. While the empirical success of SGD is often attributed to its computational efficiency and favorable generalization behavior, neither effect is well understood and disentangling them remains an open problem. Even in the simple setting of convex quadratic problems, worst-case analyses give an asymptotic convergence rate for SGD that is no better than full-batch gradient descent (GD), and the purported implicit regularization effects of SGD lack a precise explanation. In this work, we study the dynamics of multi-pass SGD on high-dimensional convex quadratics and establish an asymptotic equivalence to a stochastic differential equation, which we call homogenized stochastic gradient descent (HSGD), whose solutions we characterize explicitly in terms of a Volterra integral equation. These results yield precise formulas for the learning and risk trajectories, which reveal a mechanism of implicit conditioning that explains the efficiency of SGD relative to GD. We also prove that the noise from SGD negatively impacts generalization performance, ruling out the possibility of any type of implicit regularization in this context. Finally, we show how to adapt the HSGD formalism to include streaming SGD, which allows us to produce an exact prediction for the excess risk of multi-pass SGD relative to that of streaming SGD (bootstrap risk).
In this letter, we introduce a class of improved DC-free 4B6B codes in terms of error correction capabilities for a serially concatenated ar… (voir plus)chitecture. There are billions of different codebooks that can be derived from the 16 codewords contained in the traditional 4B6B code as per the IEEE 802.15.7 standard for visible light communication (VLC). These codebooks can be classified based on distances properties which determine their error correction performances. The traditional 4B6B code is suitable for hard-decision decoding, however, when a soft decoder is used like in a serially concatenated architecture, that code becomes obsolete. Simulations show that the proposed 4B6B code concatenated with forward error correction (FEC) codes, has better performance compared to state-of-the-art schemes such as the original 4B6B code, the enhanced Miller code, the Manchester code, the 5B10B code and the (0,4) 2/3 RLL code.
In this letter, we introduce a class of improved DC-free 4B6B codes in terms of error correction capabilities for a serially concatenated ar… (voir plus)chitecture. There are billions of different codebooks that can be derived from the 16 codewords contained in the traditional 4B6B code as per the IEEE 802.15.7 standard for visible light communication (VLC). These codebooks can be classified based on distances properties which determine their error correction performances. The traditional 4B6B code is suitable for hard-decision decoding, however, when a soft decoder is used like in a serially concatenated architecture, that code becomes obsolete. Simulations show that the proposed 4B6B code concatenated with forward error correction (FEC) codes, has better performance compared to state-of-the-art schemes such as the original 4B6B code, the enhanced Miller code, the Manchester code, the 5B10B code and the (0,4) 2/3 RLL code.
Presuppositions are assumptions that are taken for granted by an utterance, and identifying them is key to a pragmatic interpretation of lan… (voir plus)guage. In this paper, we investigate the capabilities of transformer models to perform NLI on cases involving presupposition. First, we present simple heuristics to create alternative “contrastive” test cases based on the ImpPres dataset and investigate the model performance on those test cases. Second, to better understand how the model is making its predictions, we analyze samples from sub-datasets of ImpPres and examine model performance on them. Overall, our findings suggest that NLI-trained transformer models seem to be exploiting specific structural and lexical cues as opposed to performing some kind of pragmatic reasoning.
Mutual Information (MI) has been widely used as a loss regularizer for training neural networks. This has been particularly effective when l… (voir plus)earn dis-entangled or compressed representations of high dimensional data. However, differential entropy (DE), another fundamental measure of information, has not found widespread use in neural network training. Although DE offers a potentially wider range of applications than MI, off-the-shelf DE estimators are either non differentiable, computationally intractable or fail to adapt to changes in the underlying distribution. These drawbacks prevent them from being used as regularizers in neural networks training. To address shortcomings in previously proposed estimators for DE, here we introduce K NIFE , a fully parameterized, differentiable kernel-based estimator of DE. The flexibility of our approach also allows us to construct K NIFE -based estimators for conditional (on either discrete or continuous variables) DE, as well as MI. We empirically validate our method on high-dimensional synthetic data and further apply it to guide the training of neural networks for real-world tasks. Our experiments on a large variety of tasks, including visual domain adaptation, textual fair classification, and textual fine-tuning demonstrate the effectiveness of K NIFE - based estimation. Code can be found at https: //github.com/g-pichler/knife .
Among attempts at giving a theoretical account of the success of deep neural networks, a recent line of work has identified a so-called `laz… (voir plus)y' training regime in which the network can be well approximated by its linearization around initialization. Here we investigate the comparative effect of the lazy (linear) and feature learning (non-linear) regimes on subgroups of examples based on their difficulty. Specifically, we show that easier examples are given more weight in feature learning mode, resulting in faster training compared to more difficult ones. In other words, the non-linear dynamics tends to sequentialize the learning of examples of increasing difficulty. We illustrate this phenomenon across different ways to quantify example difficulty, including c-score, label noise, and in the presence of easy-to-learn spurious correlations. Our results reveal a new understanding of how deep networks prioritize resources across example difficulty.
We propose an analysis in fair learning that preserves the utility of the data while reducing prediction disparities under the criteria of g… (voir plus)roup sufficiency. We focus on the scenario where the data contains multiple or even many subgroups, each with limited number of samples. As a result, we present a principled method for learning a fair predictor for all subgroups via formulating it as a bilevel objective. Specifically, the subgroup specific predictors are learned in the lower-level through a small amount of data and the fair predictor. In the upper-level, the fair predictor is updated to be close to all subgroup specific predictors. We further prove that such a bilevel objective can effectively control the group sufficiency and generalization error. We evaluate the proposed framework on real-world datasets. Empirical evidence suggests the consistently improved fair predictions, as well as the comparable accuracy to the baselines.