Pablo Piantanida

2023-06-06

ArXiv (prépublication)

On the incompatibility of accuracy and equal opportunity

Carlos Pinzón

Catuscia Palamidessi

Frank Valencia

2023-05-02

Machine-mediated learning (publié)

A Halfspace-Mass Depth-Based Method for Adversarial Attack Detection

Marine Picot

Federica Granese

Guillaume Staerman

Marco Romanelli

Francisco Messina

Pierre Colombo

2023-03-28

TMLR (accepté)

openreview.net

Unsupervised Layer-wise Score Aggregation for Textual OOD Detection

Maxime Darrin

Guillaume Staerman

Eduardo Dadalto Câmara Gomes

Jackie Cheung

Pierre Colombo

2023-02-20

ArXiv (prépublication)

A Minimax Approach Against Multi-Armed Adversarial Attacks Detection

Federica Granese

Marco Romanelli

Siddharth Garg

2023-02-04

ArXiv (prépublication)

Optimal Transport for Unsupervised Hallucination Detection in Neural Machine Translation

Nuno M. Guerreiro

Pierre Colombo

André Martins

Neural machine translation (NMT) has become the de-facto standard in real-world machine translation applications. However, NMT models can un… (voir plus)predictably produce severely pathological translations, known as hallucinations, that seriously undermine user trust. It becomes thus crucial to implement effective preventive strategies to guarantee their proper functioning. In this paper, we address the problem of hallucination detection in NMT by following a simple intuition: as hallucinations are detached from the source content, they exhibit encoder-decoder attention patterns that are statistically different from those of good quality translations. We frame this problem with an optimal transport formulation and propose a fully unsupervised, plug-in detector that can be used with any attention-based NMT model. Experimental results show that our detector not only outperforms all previous model-based detectors, but is also competitive with detectors that employ external models trained on millions of samples for related tasks such as quality estimation and cross-lingual sentence similarity.

2023-01-01

ACL (1) (publié)

On the (Im)Possibility of Estimating Various Notions of Differential Privacy (short paper)

Daniele Gorla

Louis Jalouzot

Federica Granese

Catuscia Palamidessi

We analyze to what extent final users can infer information about the level of protection of their data when the data obfuscation mechanism … (voir plus)is a priori unknown to them (the so-called “black-box" scenario). In particular, we delve into the investigation of two notions of local differential privacy (LDP), namely 𝜀 -LDP and Rényi LDP. On one hand, we prove that, without any assumption on the underlying distributions, it is not possible to have an algorithm able to infer the level of data protection with provable guarantees. On the other hand, we demonstrate that, under reasonable assumptions (namely, Lipschitzness of the involved densities on a closed interval), such guarantees exist and can be achieved by a simple histogram-based estimator.

2023-01-01

ICTCS (publié)

dblp.uni-trier.de

Beyond Mahalanobis-Based Scores for Textual OOD Detection

Pierre Colombo

Eduardo Dadalto Câmara Gomes

Guillaume Staerman

Nathan Noiry

2022-11-24

ArXiv (prépublication)

Beyond Mahalanobis Distance for Textual OOD Detection

Pierre Colombo

Eduardo Dadalto Câmara Gomes

Guillaume Staerman

Nathan Noiry

openreview.net

KNIFE: Kernelized-Neural Differential Entropy Estimation

Georg Pichler

Pierre Colombo

Malik Boudiaf

Gunther Koliander

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 .

2022-01-01

arXiv.org (prépublication)

dblp.uni-trier.de

Realistic Evaluation of Transductive Few-Shot Learning - Supplementary Material

Olivier Veilleux

Éts Montréal

Malik Boudiaf

Ismail Ben

Ayed Éts Montreal

In the main tables of the paper, we did not include the performances of α-TIM in the standard balanced setting. Here, we emphasize that α-… (voir plus)TIM is a generalization of TIM [1] as when α → 1 (i.e., the α-entropies tend to the Shannon entropies), α-TIM tends to TIM. Therefore, in the standard setting, where optimal hyper-parameter α is obtained over validation tasks that are balanced (as in the standard validation tasks of the original TIM and the other existing methods), the performance of α-TIM is the same as TIM. When α is tuned on balanced validation tasks, we obtain an optimal value of α very close to 1, and our α-mutual information approaches the standard mutual information. When the validation tasks are uniformly random, as in our new setting and in the validation plots we provided in the main figure, one can see that the performance of α-TIM remains competitive when we tend to balanced testing tasks (i.e., when a is increasing), but is significantly better than TIM when we tend to uniformly-random testing tasks (a = 1). These results illustrate the flexibility of α-divergences, and are in line with the technical analysis provided in the main paper.

Learning Anonymized Representations with Adversarial Neural Networks

Clément Feutry

Yoshua Bengio

P. Duhamel

Statistical methods protecting sensitive information or the identity of the data owner have become critical to ensure privacy of individuals… (voir plus) as well as of organizations. This paper investigates anonymization methods based on representation learning and deep neural networks, and motivated by novel information theoretical bounds. We introduce a novel training objective for simultaneously training a predictor over target variables of interest (the regular labels) while preventing an intermediate representation to be predictive of the private labels. The architecture is based on three sub-networks: one going from input to representation, one from representation to predicted regular labels, and one from representation to predicted private labels. The training procedure aims at learning representations that preserve the relevant part of the information (about regular labels) while dismissing information about the private labels which correspond to the identity of a person. We demonstrate the success of this approach for two distinct classification versus anonymization tasks (handwritten digits and sentiment analysis).

2018-02-26

ArXiv (prépublication)