Adversarial Logit Pairing shows 0.6% accuracy in threat model.
problem Evaluating robustness of Adversarial Logit Pairing against adversarial examples.
method Evaluation of Adversarial Logit Pairing's robustness through accuracy testing.
result Adversarial Logit Pairing achieves 0.6% accuracy in threat model.
In this paper, we develop improved techniques for defending against adversarial examples at scale. First, we implement the state of the art version of adversarial training at unprecedented scale on ImageNet and investigate whether it remains effective in this setting - an important open scientific question (Athalye et …
Logit pairing methods don't provide robustness against adversarial examples.
problem Improving adversarial robustness of classifiers.
method Logit regularization methods (CLP, LSQ, ALP) to craft adversarial examples.
result Adversarial Logit Pairing (ALP) may provide robustness, but the increase is small.
Improves adversarial robustness by constraining logits with a bounded function.
problem Improving adversarial robustness in deep learning models.
method Addition of a bounded function before softmax to constrain logits.
result Our method improves adversarial robustness without requiring adversarial training.
Proposes SOVR loss to improve adversarial robustness by increasing logit margins.
problem Adversarial training's difficulty in robustness against sophisticated attacks.
method Introduces SOVR loss function that switches from cross-entropy to one-vs-the-rest loss for important samples.
result SOVR loss increases logit margins of important samples, improving robustness against Auto-Attack.
A two-layer network corrects logits to defend adversarial attacks.
problem Detecting and defending adversarial attacks on deep learning models.
method A two-layer network trained on mixed logits to recover original predictions.
result Shows promising results in defense with high accuracy and interpretability.
New method bypasses statistical and classifier-based detection of adversarial examples.
problem Vulnerability of deep learning classifiers to adversarial examples.
method Classifier-based adaptation of statistical test method and Logit Mimicry Attack.
result Reduces detection performance to less than 2.2% TPR and 1.6% TPR for statistical test and classifier-based methods, respectively, even at 5% FPR.
Method identifies over-optimized adversarial examples using IQR-based logit thresholding.
problem Over-optimized adversarial examples are hard to detect due to model optimization.
method IQR-based logit thresholding method for identifying over-optimized adversarial examples.
result The logits of the model are more useful in detecting over-optimized adversarial examples.
Simple regularization methods mimic adversarial training's robustness.
problem Expensive adversarial training for robustness.
method Label smoothing and logit squeezing.
result Achieves strong adversarial robustness without adversarial examples.
The paper proposes logit regularization for improving neural network robustness against adversarial attacks.
problem Vulnerability of neural networks to adversarial examples, especially in safety-critical applications.
method Logit regularization techniques combined with other methods to enhance adversarial robustness.
result Logit regularization can improve the effectiveness of existing adversarial defenses and create stronger black-box attacks.
Generative classifier derived from any discriminative classifier rejects illegal inputs.
problem Detecting and rejecting illegal inputs like adversarial examples and out-of-distribution samples.
method SDIM-logit: learns generative classifier from logits of any discriminative classifier, imposing statistical constraints.
result SDIM-logit inherits performance of base classifier without loss and can reject illegal inputs.
Detects misclassifications and adversarial examples using neural network logits.
problem Unable to detect misclassifications and adversarial examples in neural networks.
method Introspection using pretrained neural network logits.
result Simple 3-layer neural network trained on logits detects misclassifications competitively.
LAWN normalizes logits to improve deep network adaptability and generalization.
problem Large logits and weights lead to overfitting in deep networks.
method Logit Attenuating Weight Normalization (LAWN) constrains weight norms in the final sub-network.
result LAWN improves generalization and adaptability of deep networks.
This work certifies non-uniform bounds against adversarial attacks for neural networks.
problem Certifying robust regions around data points against non-uniform adversarial attacks.
method Formulated as an optimization problem with nonlinear constraints, using the augmented Lagrangian method for general feedforward neural networks.
result Non-uniform bounds have larger volumes and better interpretability compared to uniform bounds.
New method distills cloud models into edge-friendly ones.
problem Cloud-to-edge model compression with limited data exchange.
method Two-step workflow of deprivatization and distillation.
result Outperforms previous state-of-the-art approaches on various benchmarks.
The paper connects discrete choice models to multi-armed bandit algorithms with sublinear regret bounds.
problem Optimizing user choices in a multi-armed bandit setting.
method Establishes connections between discrete choice models and multi-armed bandit algorithms, providing sublinear regret bounds and novel algorithms.
result Sublinear regret bounds for a family of algorithms, including the Exp3 algorithm.
Adversarial training linked to operator norm regularization, proving network sensitivity to attacks.
problem Robustifying neural networks against adversarial attacks.
method Theoretical link established between adversarial training and operator norm regularization.
result Adversarial training is equivalent to data-dependent operator norm regularization.
Logit regularization induces logit clustering, affecting classifier performance.
problem Understanding the mechanism of logit regularization in classification.
method Analysis of logit regularization in linear classification, proving logit clustering leads to Fisher's Linear Discriminant alignment.
result Logit regularization can halve critical sample complexity and induce robust generalization.
A neural network-based two-sample test improves classification accuracy.
problem Differentiating between two sub-exponential densities.
method Difference of logit function from trained classification neural network.
result Network complexity scales with intrinsic dimensionality for low-dimensional manifolds.
Logit-GFN accelerates GFlowNets training by scaling logits based on temperature.
problem Training temperature-conditional GFlowNets is numerically challenging.
method Logit-GFN uses a learned function of temperature to scale policy logits.
result Logit-GFN greatly accelerates GFlowNets training and improves generalization and mode discovery.
A new logit model derived from the Weibull manifold.
problem No potential function on the Weibull manifold.
method Extracted a logit model from the two-parameter Weibull model.
result Found a completely integrable Hamiltonian gradient system on the logit model.
Logit-Coordinate models improve representation of mixed data types.
problem Representation of mixed continuous-categorical data in generative models.
method Logit-coordinate framework combining categorical and continuous variables.
result Logit-Coordinate models outperform one-hot encoding in simulations and real data.
Linear filtration helps delete training data from models.
problem Deleting training data from models when individuals request it.
method Linear filtration as a computationally efficient sanitization method.
result Demonstrates benefits in an adversarial setting over naive deletion schemes.
New global adversarial attacks improve DNN robustness assessment.
problem Vulnerability of deep neural networks to adversarial attacks.
method Proposed global adversarial example pairs and attack methods.
result DNNs hardened with local adversarial training are vulnerable to global attacks.
Study analyzes overfitting in neural nets under class imbalance, proposing new loss functions.
problem Overfitting and class imbalance in neural nets, particularly in image segmentation.
method Analyzes logits distribution, derives asymmetric loss functions and regularizers.
result Proposed loss functions improve segmentation performance in brain tumor core.
Motivated by generating personalized recommendations using ordinal (or preference) data, we study the question of learning a mixture of MultiNomial Logit (MNL) model, a parameterized class of distributions over permutations, from partial ordinal or preference data (e.g. pair-wise comparisons). Despite its long standing…
The paper proposes effective margin regularization to improve adversarial robustness in deep neural networks.
problem Adversarial vulnerability of deep neural networks (DNNs).
method Regularization of effective weight norm during training to maximize effective margins.
result Effective margin regularization (EMR) boosts adversarial robustness in both standard and adversarial training.
New method tightens federated probe-logit distillation rates under varying bandwidths.
problem Estimating conditional distributions in federated learning with heterogeneous bandwidth constraints.
method Developed a new federated probe-logit distillation (FPLD) method with optimal allocation for varying bandwidths.
result Achieved matching lower and upper bounds for the minimax rate under heterogeneous bandwidths.
In the past decades, intensive efforts have been put to design various loss functions and metric forms for metric learning problem. These improvements have shown promising results when the test data is similar to the training data. However, the trained models often fail to produce reliable distances on the ambiguous te…
ManiGen generates adversarial examples without classifier knowledge.
problem Vulnerability of neural network classifiers to adversarial examples.
method Generates adversarial examples by searching along the manifold.
result Adversarial examples generated by ManiGen are as successful as state-of-the-art generators.
This paper compares machine learning and logit models for predicting travel mode choice.
problem Comparing interpretability of machine learning and logit models for travel mode choice.
method Empirical evaluation of machine learning and logit models on stated preference survey data.
result Machine learning produces significantly higher predictive accuracy than logit models.
It is becoming increasingly clear that many machine learning classifiers are vulnerable to adversarial examples. In attempting to explain the origin of adversarial examples, previous studies have typically focused on the fact that neural networks operate on high dimensional data, they overfit, or they are too linear. H…
Annealing Double-Head calibrates deep neural networks during training.
problem Overestimation or underestimation of predictive confidence in deep neural networks.
method An additional calibration head and Annealing technique to dynamically scale logits.
result State-of-the-art model calibration performance achieved without post-processing.
Logit dynamics formula reveals self-regulation in softmax policy gradient methods.
problem Understanding the stability and convergence of softmax policy gradient methods.
method Deriving the exact formula for the L2 norm of the logit update vector.
result Logit update magnitudes are modulated by action probability and policy concentration.
Proposes a convex model for mixed logit to handle individual heterogeneity.
problem Non-convex optimization in mixed logit models for individual heterogeneity.
method Sparse and low-rank decomposition for convex formulation.
result Convex formulation avoids simulation-based approximation and unstable model interpretation.
PLD distills knowledge using choice-theoretic Plackett-Luce model.
problem Model compression and knowledge transfer between large and small networks.
method PLD uses a weighted list-wise ranking loss based on the Plackett-Luce model.
result PLD achieves consistent gains across diverse architectures and distillation methods.
Study dynamic assortment and positioning of products with varying display effects.
problem Dynamic assortment and positioning of products with varying display effects.
method Design round-based learning algorithms for both multiplicative and general position effects models, and develop efficient subroutines for optimization.
result First regret-optimal characterization for both models, with matching upper and lower bounds.
MANO normalizes logits to estimate test accuracy without labels.
problem Estimating test accuracy of OOD samples without labels.
method Applies Lp norm to normalized logits. result Achieves state-of-the-art performance across various architectures.
New method uses low logit rank to simplify complex language models.
problem Understanding and learning from modern language models.
method Exploiting the low logit rank structure of language models for efficient learning.
result An efficient algorithm for learning low logit rank models from queries.
Logit distance bounds representational similarity of models.
problem Approximating linear similarity when distributions are close.
method Defined a logit distance and proved its relationship to representational dissimilarity.
result Logit distance bounds representational similarity, providing nontrivial control in practice.
New method lures adversaries to choose false directions in black-box attacks.
problem Improving robustness against black-box transfer attacks.
method Incorporates a removable neural network to induce a luring effect.
result Demonstrates effectiveness in thwarting state-of-the-art attacks.
New models improve choice prediction accuracy.
problem Model misspecifications in discrete choice models lead to limited predictability and biased estimates.
method Proposes a new approach to estimate choice models by dividing the systematic part into knowledge-driven and data-driven components, which learns a new representation from available variables.
result The new models (L-MNL and L-NL) outperform traditional models in predictive performance and parameter estimation.
Improved unsupervised word translation using adversarial autoencoder with cycle consistency and input reconstruction.
problem Challenging language pairs and lack of parallel data for unsupervised word translation.
method Adversarial autoencoder with cycle consistency and input reconstruction regularization.
result More stable and better performance than recent approaches.
Motivated by the problem of automated repair of software vulnerabilities, we propose an adversarial learning approach that maps from one discrete source domain to another target domain without requiring paired labeled examples or source and target domains to be bijections. We demonstrate that the proposed adversarial l…
A Triangle Generative Adversarial Network (Δ-GAN) is developed for semi-supervised cross-domain joint distribution matching, where the training data consists of samples from each domain, and supervision of domain correspondence is provided by only a few paired samples. Δ-GAN consists of four neural networks, two ge…
We consider neural network training, in applications in which there are many possible classes, but at test-time, the task is a binary classification task of determining whether the given example belongs to a specific class, where the class of interest can be different each time the classifier is applied. For instance, …
New algorithm reduces switching costs in multinomial logit bandit problems.
problem Minimizing switching costs in multinomial logit bandit problems.
method Proposed AT-DUCB and FH-DUCB algorithms with low assortment switching costs.
result AT-DUCB and FH-DUCB algorithms achieve almost optimal minimax regret with low switching costs.
New method uses adversarial training for blind source separation.
problem Blind single-channel source separation challenge.
method Adversarial training for independence of sources.
result Good performance validated on image sources.