Search Results for author:
Found 59 papers, 25 papers with code
Is Meta-training Really Necessary for Molecular Few-Shot Learning ?
Few-shot learning has recently attracted significant interest in drug discovery, with a recent, fast-growing literature mostly involving convoluted meta-learning strategies.
$\texttt{COSMIC}$: Mutual Information for Task-Agnostic Summarization Evaluation
Assessing the quality of summarizers poses significant challenges.
Optimal Zero-Shot Detector for Multi-Armed Attacks
We approach this defensive strategy with utmost caution, operating in an environment where the defender possesses significantly less information compared to the attacker.
On the Convergence of Semi Unsupervised Calibration through Prior Adaptation Algorithm
The map derived by this system} has the peculiarity of being non-hyperbolic {with a non-bounded set of non-isolated fixed points}.
Preserving Privacy in GANs Against Membership Inference Attack
In the present work, the overfitting in GANs is studied in terms of the discriminator, and a more general measure of overfitting based on the Bhattacharyya coefficient is defined.
Transductive Learning for Textual Few-Shot Classification in API-based Embedding Models
First, we introduce a scenario where the embedding of a pre-trained model is served through a gated API with compute-cost and data-privacy constraints.
Toward Stronger Textual Attack Detectors
The landscape of available textual adversarial attacks keeps growing, posing severe threats and raising concerns regarding the deep NLP system's integrity.
Fundamental Limits of Membership Inference Attacks on Machine Learning Models
Membership inference attacks (MIA) can reveal whether a particular data point was part of the training dataset, potentially exposing sensitive information about individuals.
Unsupervised Calibration through Prior Adaptation for Text Classification using Large Language Models
These models are usually trained with a very large amount of unsupervised text data and adapted to perform a downstream natural language task using methods like fine-tuning, calibration or in-context learning.
A Functional Data Perspective and Baseline On Multi-Layer Out-of-Distribution Detection
A key feature of out-of-distribution (OOD) detection is to exploit a trained neural network by extracting statistical patterns and relationships through the multi-layer classifier to detect shifts in the expected input data distribution.
A Data-Driven Measure of Relative Uncertainty for Misclassification Detection
Misclassification detection is an important problem in machine learning, as it allows for the identification of instances where the model's predictions are unreliable.
Unsupervised Layer-wise Score Aggregation for Textual OOD Detection
More importantly, we show that the usual choice (the last layer) is rarely the best one for OOD detection and that far better results could be achieved if the best layer were picked.
A Minimax Approach Against Multi-Armed Adversarial Attacks Detection
Multi-armed adversarial attacks, in which multiple algorithms and objective loss functions are simultaneously used at evaluation time, have been shown to be highly successful in fooling state-of-the-art adversarial examples detectors while requiring no specific side information about the detection mechanism.
Open-Set Likelihood Maximization for Few-Shot Learning
We tackle the Few-Shot Open-Set Recognition (FSOSR) problem, i. e. classifying instances among a set of classes for which we only have a few labeled samples, while simultaneously detecting instances that do not belong to any known class.
Optimal Transport for Unsupervised Hallucination Detection in Neural Machine Translation
Neural machine translation (NMT) has become the de-facto standard in real-world machine translation applications.
Rainproof: An Umbrella To Shield Text Generators From Out-Of-Distribution Data
In this work, we focus on leveraging soft-probabilities in a black-box framework, i. e. we can access the soft-predictions but not the internal states of the model.
Beyond Mahalanobis-Based Scores for Textual OOD Detection
Deep learning methods have boosted the adoption of NLP systems in real-life applications.
The Glass Ceiling of Automatic Evaluation in Natural Language Generation
Automatic evaluation metrics capable of replacing human judgments are critical to allowing fast development of new methods.
MEAD: A Multi-Armed Approach for Evaluation of Adversarial Examples Detectors
Detection of adversarial examples has been a hot topic in the last years due to its importance for safely deploying machine learning algorithms in critical applications.
Model-Agnostic Few-Shot Open-Set Recognition
Through extensive experiments spanning 5 datasets, we show that OSTIM surpasses both inductive and existing transductive methods in detecting open-set instances while competing with the strongest transductive methods in classifying closed-set instances.
Learning Disentangled Textual Representations via Statistical Measures of Similarity
When working with textual data, a natural application of disentangled representations is fair classification where the goal is to make predictions without being biased (or influenced) by sensitive attributes that may be present in the data (e. g., age, gender or race).
Realistic Evaluation of Transductive Few-Shot Learning
Transductive inference is widely used in few-shot learning, as it leverages the statistics of the unlabeled query set of a few-shot task, typically yielding substantially better performances than its inductive counterpart.
Perfectly Accurate Membership Inference by a Dishonest Central Server in Federated Learning
Federated Learning is expected to provide strong privacy guarantees, as only gradients or model parameters but no plain text training data is ever exchanged either between the clients or between the clients and the central server.
Leveraging Adversarial Examples to Quantify Membership Information Leakage
The use of personal data for training machine learning systems comes with a privacy threat and measuring the level of privacy of a model is one of the major challenges in machine learning today.
Igeood: An Information Geometry Approach to Out-of-Distribution Detection
Reliable out-of-distribution (OOD) detection is fundamental to implementing safer modern machine learning (ML) systems.
Out-of-Distribution Detection
Out of Distribution (OOD) Detection
A Differential Entropy Estimator for Training Neural Networks
Mutual Information (MI) has been widely used as a loss regularizer for training neural networks.
PACMAN: PAC-style bounds accounting for the Mismatch between Accuracy and Negative log-loss
In this work, we introduce an analysis based on point-wise PAC approach over the generalization gap considering the mismatch of testing based on the accuracy metric and training on the negative log-loss.
InfoLM: A New Metric to Evaluate Summarization & Data2Text Generation
In this paper, we introduce InfoLM a family of untrained metrics that can be viewed as a string-based metric that addresses the aforementioned flaws thanks to a pre-trained masked language model.
Automatic Text Evaluation through the Lens of Wasserstein Barycenters
A new metric \texttt{BaryScore} to evaluate text generation based on deep contextualized embeddings e. g., BERT, Roberta, ELMo) is introduced.
Learning Sparse Privacy-Preserving Representations for Smart Meters Data
We formulate this as the problem of learning a sparse representation of SMs data with minimum information leakage and maximum utility.
On the impossibility of non-trivial accuracy under fairness constraints
One of the main concerns about fairness in machine learning (ML) is that, in order to achieve it, one may have to trade off some accuracy.
Mutual-Information Based Few-Shot Classification
We motivate our transductive loss by deriving a formal relation between the classification accuracy and mutual-information maximization.
Adversarial Robustness via Fisher-Rao Regularization
Adversarial robustness has become a topic of growing interest in machine learning since it was observed that neural networks tend to be brittle.
DOCTOR: A Simple Method for Detecting Misclassification Errors
Deep neural networks (DNNs) have shown to perform very well on large scale object recognition problems and lead to widespread use for real-world applications, including situations where DNN are implemented as "black boxes".
Bounding Information Leakage in Machine Learning
We present a novel formalism, generalizing membership and attribute inference attack setups previously studied in the literature and connecting them to memorization and generalization.
A Novel Estimator of Mutual Information for Learning to Disentangle Textual Representations
We show the superiority of this method on fair classification and on textual style transfer tasks.
Ranked #1 on
Text Style Transfer
on Yelp Review Dataset (Large)
(BLEU metric)
Learning to Disentangle Textual Representations and Attributes via Mutual Information
Learning disentangled representations of textual data is essential for many natural language tasks such as fair classification (\textit{e. g.} building classifiers whose decisions cannot disproportionately hurt or benefit specific groups identified by sensitive attributes), style transfer and sentence generation, among others.
Few-Shot Segmentation Without Meta-Learning: A Good Transductive Inference Is All You Need?
We show that the way inference is performed in few-shot segmentation tasks has a substantial effect on performances -- an aspect often overlooked in the literature in favor of the meta-learning paradigm.
Ranked #3 on
Few-Shot Semantic Segmentation
on COCO-20i (10-shot)
Privacy-Preserving Synthetic Smart Meters Data
However, the use of power consumption data raises significant privacy concerns, as this data usually belongs to clients of a power company.
Information Maximization for Few-Shot Learning
We introduce Transductive Infomation Maximization (TIM) for few-shot learning.
Deep Directed Information-Based Learning for Privacy-Preserving Smart Meter Data Release
In this paper, we study this problem in the context of time series data and smart meters (SMs) power consumption measurements in particular.
The Role of Mutual Information in Variational Classifiers
In practice, this behaviour is controlled by various--sometimes heuristics--regularization techniques, which are motivated by developing upper bounds to the generalization error.
Transductive Information Maximization For Few-Shot Learning
We introduce Transductive Infomation Maximization (TIM) for few-shot learning.
On the Impact of Side Information on Smart Meter Privacy-Preserving Methods
On the one hand, the releaser in the CAL method, by getting supervision from the actual values of the private variables and feedback from the adversary performance, tries to minimize the adversary log-likelihood.
Privacy-Cost Management in Smart Meters with Mutual Information-Based Reinforcement Learning
Unlike previous studies, we model the whole temporal correlation in the data to learn the MI in its general form and use a neural network to estimate the MI-based reward signal to guide the PCMU learning process.
Estimating g-Leakage via Machine Learning
A feature of our approach is that it does not require to estimate the conditional probabilities, and that it is suitable for a large class of ML algorithms.
A unifying mutual information view of metric learning: cross-entropy vs. pairwise losses
Second, we show that, more generally, minimizing the cross-entropy is actually equivalent to maximizing the mutual information, to which we connect several well-known pairwise losses.
Ranked #12 on
Metric Learning
on CARS196
(using extra training data)
Privacy-Cost Management in Smart Meters Using Deep Reinforcement Learning
Smart meters (SMs) play a pivotal rule in the smart grid by being able to report the electricity usage of consumers to the utility provider (UP) almost in real-time.
On the Estimation of Information Measures of Continuous Distributions
In particular, we provide confidence bounds for simple histogram based estimation of differential entropy from a fixed number of samples, assuming that the probability density function is Lipschitz continuous with known Lipschitz constant and known, bounded support.
Real-Time Privacy-Preserving Data Release for Smart Meters
In this paper, we focus on real-time privacy threats, i. e., potential attackers that try to infer sensitive information from SMs data in an online fashion.
Understanding the Behaviour of the Empirical Cross-Entropy Beyond the Training Distribution
Our main result is that the testing gap between the empirical cross-entropy and its statistical expectation (measured with respect to the testing probability law) can be bounded with high probability by the mutual information between the input testing samples and the corresponding representations, generated by the encoder obtained at training time.
On Direct Distribution Matching for Adapting Segmentation Networks
We juxtapose our approach to state-of-the-art segmentation adaptation via adversarial training in the network-output space.
Learning Anonymized Representations with Adversarial Neural Networks
Statistical methods protecting sensitive information or the identity of the data owner have become critical to ensure privacy of individuals as well as of organizations.
The Role of Information Complexity and Randomization in Representation Learning
This paper presents a sample-dependent bound on the generalization gap of the cross-entropy loss that scales with the information complexity (IC) of the representations, meaning the mutual information between inputs and their representations.
Compression-Based Regularization with an Application to Multi-Task Learning
This paper investigates, from information theoretic grounds, a learning problem based on the principle that any regularity in a given dataset can be exploited to extract compact features from data, i. e., using fewer bits than needed to fully describe the data itself, in order to build meaningful representations of a relevant content (multiple labels).
The Multi-layer Information Bottleneck Problem
Based on information forwarded by the preceding layer, each stage of the network is required to preserve a certain level of relevance with regards to a specific hidden variable, quantified by the mutual information.
Collaborative Information Bottleneck
On the other hand, in CDIB there are two cooperating encoders which separately observe $X_1$ and $X_2$ and a third node which can listen to the exchanges between the two encoders in order to obtain information about a hidden variable $Y$.
Distributed Information-Theoretic Clustering
We study a novel multi-terminal source coding setup motivated by the biclustering problem.
