Method generates time-series attribution maps with identifiability guarantees.
problem Lack of identifiability guarantees in gradient-based attribution methods.
method Regularized contrastive learning algorithm trained on time-series data.
result Empirically shows robust approximation of zero vs. non-zero entries in the ground-truth attribution map.
Adapts IG for better feature attributions and robustness.
problem Reliability concerns in feature attributions for deep learning models.
method Adaptation of path-based feature attribution to Riemannian geometry of data manifolds.
result IG along geodesics generates more intuitive and robust explanations.
RIG extends IG to Riemannian manifolds for explainable AI.
problem Lack of explainability in AI models.
method Extension of Integrated Gradients to Riemannian manifolds.
result RIG restricts to IG in Euclidean space.
Recent research has demonstrated that feature attribution methods for deep networks can themselves be incorporated into training; these attribution priors optimize for a model whose attributions have certain desirable properties -- most frequently, that particular features are important or unimportant. These attributio…
Proposes new attribution methods for trees with regularization.
problem Feature attribution for trees trained with regularization.
method Prediction Decomposition Attribution (PreDecomp) and TreeInner.
result TreeInner shows state-of-the-art feature selection performance.
Greedy PIG adapts integrated gradients for better feature attribution.
problem Interpreting deep learning model predictions.
method Unified discrete optimization framework for feature attribution and selection.
result Greedy PIG improves feature attribution on various tasks.
Improved saliency maps for deep neural networks with reduced noise.
problem Noisy explanations in Integrated Gradients for deep neural networks.
method SmoothTaylor, adaptive noising, and SmoothGrad techniques.
result SmoothTaylor and adaptive noising generate better quality saliency maps.
Proposes Geodesic Integrated Gradients (GIG) for more accurate feature attributions in deep networks.
problem Flawed attributions using straight paths from Integrated Gradients (IG).
method Introduces a model-induced Riemannian metric and computes attributions along geodesics.
result GIG produces more faithful attributions than IG on benchmarks.
Integrated gradients are widely employed to evaluate the contribution of input features in classification models because it satisfies the axioms for attribution of prediction. This method, however, requires an appropriate baseline for reliable determination of the contributions. We propose a compensated integrated grad…
PS-IG improves feature attribution by reducing noise and variance.
problem Improving feature attribution in machine learning models.
method Path-sampled integrated gradients (PS-IG) computes expected value over sampled baselines.
result PS-IG reduces attribution variance by a factor of 1/3 under uniform sampling.
Enhanced visual feature attribution via adaptive baseline weighting.
problem IG's sensitivity to baseline images leads to noisy or unstable explanations.
method Weighted Integrated Gradients (WG) evaluates and weights baselines for improved reliability.
result WG improves over Expected Gradients (EG) by up to 36% across various models.
MACQ method explains deep learning models by analyzing feature contributions across prediction levels.
problem Explaining deep learning model predictions.
method Global gradient-based, model-agnostic approach focusing on marginal attribution.
result MACQ separates feature contributions from interaction effects and visualizes 3-way relationships.
We develop a theory of higher-order feature attribution for complex models.
problem Interpreting feature contributions in models with interactions is challenging.
method We extend Integrated Gradients (IG) to higher-order feature attributions.
result We establish natural connections to statistics and topological signal processing.
WassersteinGrad improves weather forecasting explanations by addressing geometric misalignment issues.
problem Improving explainability of autoregressive neural predictions on dynamic physical fields.
method WassersteinGrad, a geometric consensus method for averaged perturbed attribution maps.
result WassersteinGrad provides more accurate explanations for weather forecasting models.
Understanding the flow of information in Deep Neural Networks (DNNs) is a challenging problem that has gain increasing attention over the last few years. While several methods have been proposed to explain network predictions, there have been only a few attempts to compare them from a theoretical perspective. What is m…
Nonparametric method measures influence of training images on diffusion model outputs.
problem Quantifying influence of individual training examples on diffusion model outputs.
method Patch-level similarity between generated and training images, using optimal score function.
result Strong attribution performance, matching gradient-based approaches and outperforming baselines.
PathBoost boosts graph-level predictions using path-based features.
problem Graph-level classification and regression challenges.
method Gradient tree boosting method for graph-level prediction.
result PathBoost outperforms graph neural networks and graph kernel approaches in many cases.
New method makes neural network explanations more robust to attacks.
problem Vulnerability of feature attributions to adversarial attacks.
method Promotes Lipschitz continuity and smoothness in DNNs to improve robustness.
result Regularization and stochastic smoothing methods enhance attribution robustness.
TANGOS improves neural network performance on tabular data by encouraging neuron specialization.
problem Improving neural network performance on tabular data.
method Gradient orthogonality and specialization of latent units.
result TANGOS leads to improved out-of-sample generalization performance.
Unified analysis of removal-based feature attributions robustness.
problem Robustness of removal-based feature attributions is not well understood.
method Theoretical analysis and upper bounds derivation for removal-based feature attributions under input and model perturbations.
result Upper bounds for the difference between intact and perturbed attributions derived under various perturbation settings.
New method for efficient sketching of gradients and Hessians.
problem Memory constraints in training machine learning models.
method A novel framework for scalable gradient and HVP sketching tailored for modern hardware.
result Theoretical guarantees and practical applications in training data attribution and Hessian spectrum analysis.
Wavelet Attribution Method (WAM) improves feature attribution for deep models.
problem Inability of pixel-based heatmaps to capture data structure and variability in feature attribution.
method Wavelet domain for feature attribution, leveraging spatial and scale-localized properties of wavelet coefficients.
result WAM provides quantitatively superior explanations across audio, image, and volume modalities.
Algorithm ensures demographic parity in regression without sensitive attribute data.
problem Performing regression with demographic parity constraints.
method Post-processing algorithm using accurate estimates and sensitive attribute predictor.
result Generates predictions meeting demographic parity constraint.
Proposes a method to quantify and explain deep learning model uncertainties.
problem Deep learning model predictions are sensitive to perturbations and adversarial attacks.
method Gradient-based uncertainty attribution method to identify problematic regions and propose mitigation strategies.
result Proposed UA-Backprop method achieves competitive accuracy and efficiency compared to existing methods.
Paper estimates differences in multi-attribute Gaussian graphical models using non-convex penalties.
problem Estimating differences in multi-attribute Gaussian graphical models with similar structure.
method Penalized D-trace loss function with non-convex (log-sum and SCAD) penalties, proximal gradient descent methods.
result Theoretical analysis and numerical examples support consistency in support recovery and estimation.
ASTRA improves TDA by more accurately approximating iHVP.
problem Improving insights into training data attribution.
method ASTRA uses EKFAC-preconditioner on Neumann series iterations to accurately approximate iHVP.
result Improving iHVP approximation significantly improves TDA performance.
Attribution methods have been developed to explain the decision of a machine learning model on a given input. We use the Integrated Gradient method for finding attributions to define the causal neighborhood of an input by incrementally masking high attribution features. We study the robustness of machine learning model…
SLOGAN improves GANs' conditional generation by balancing latent attribute distributions.
problem Imbalanced latent attribute distributions in GANs.
method Stein latent optimization with a Gaussian mixture prior and contrastive loss.
result SLOGAN achieves state-of-the-art unsupervised conditional generation performance.
Paper proposes a new method to measure model sensitivity using final model only.
problem Understanding model behavior using only the final trained model.
method Reframe TDA as measuring sensitivity, propose further training as gold standard, unify gradient-based methods.
result Gradient-based methods approximate further training but vary in quality.
Study examines how machine learning attribution methods reflect risk in finance.
problem Ensuring machine learning attribution methods accurately reflect underlying risks in finance.
method Examined Shapley value and Integrated Gradients, and derived axioms from asset pricing domain knowledge.
result Neither Shapley value nor Integrated Gradients can satisfy all axioms for reflecting risks accurately.
Local explanation methods, also known as attribution methods, attribute a deep network's prediction to its input (cf. Baehrens et al. (2010)). We respond to the claim from Adebayo et al. (2018) that local explanation methods lack sensitivity, i.e., DNNs with randomly-initialized weights produce explanations that are bo…
Unified framework for linear attribution methods in deep learning.
problem Separate theoretical foundations of XAI attribution methods.
method GRALIS (Gradient-Riesz Averaged Locally-Integrated Shapley) framework.
result Unified representation theory for linear attribution methods.
Unified framework analyzes privacy risks from gradients in distributed learning.
problem Analyzing inference privacy risks from gradients in machine learning.
method Unified game-based framework for various attacks, including attribute, property, distributional, and user disclosures.
result Demonstrates inefficacy of data aggregation for privacy against inference attacks.
An emerging problem in trustworthy machine learning is to train models that produce robust interpretations for their predictions. We take a step towards solving this problem through the lens of axiomatic attribution of neural networks. Our theory is grounded in the recent work, Integrated Gradients (IG), in axiomatical…
The study compares different game-theoretic attribution methods and finds that interventional Shapley values yield less consistent results than Aumann-Shapley due to path symmetry.
problem Investigating the influence of path choice on game-theoretic attribution algorithms.
method Comparative analysis of interventional Shapley values and Generalized Integrated Gradients (GIG) methods.
result Interventional Shapley values yield less consistent attributions than Aumann-Shapley due to path symmetry and extended away from the training data manifold.
In this paper, we study the problem of transfer learning with the attribute data. In the transfer learning problem, we want to leverage the data of the auxiliary and the target domains to build an effective model for the classification problem in the target domain. Meanwhile, the attributes are naturally stable cross d…
Saliency methods can aid understanding of deep neural networks. Recent years have witnessed many improvements to saliency methods, as well as new ways for evaluating them. In this paper, we 1) present a novel region-based attribution method, XRAI, that builds upon integrated gradients (Sundararajan et al. 2017), 2) int…
Method identifies root causes of anomalies in causal processes.
problem Identifying root causes of anomalies in causal processes.
method Noisy functional causal model, Bayesian learning, gradient-based attribution.
result Proposes efficient method to compute anomaly attribution scores.
New attack recovers user-level information from large batch images.
problem Recovering private information from user-level gradients in distributed learning.
method Proposes a gradient inversion attack using a denoising diffusion model as a prior.
result Demonstrates recovery of realistic facial images and private attributes.
Interpreting predictions from tree ensemble methods such as gradient boosting machines and random forests is important, yet feature attribution for trees is often heuristic and not individualized for each prediction. Here we show that popular feature attribution methods are inconsistent, meaning they can lower a featur…
We develop new stochastic gradient methods for efficiently solving sparse linear regression in a partial attribute observation setting, where learners are only allowed to observe a fixed number of actively chosen attributes per example at training and prediction times. It is shown that the methods achieve essentially a…
RUMBoost combines RUMs and deep learning for better choice modelling.
problem Creating interpretable and robust discrete choice models.
method Gradient Boosted Regression Trees for utility functions, with constraints for interpretability and monotonicity.
result RUMBoost outperforms ML and RUM benchmarks in predictive performance and interpretability.
This research compares two encoding methods for categorical attributes in machine learning, affecting model fairness.
problem The impact of encoding protected categorical attributes on fairness in machine learning models.
method Comparison of one-hot encoding and target encoding methods.
result Target encoding can lead to more unfair models compared to one-hot encoding due to induced bias.
Deep generative neural networks have proven effective at both conditional and unconditional modeling of complex data distributions. Conditional generation enables interactive control, but creating new controls often requires expensive retraining. In this paper, we develop a method to condition generation without retrai…
New framework improves attribution of predictive uncertainties in classification models.
problem Improper attribution of predictive uncertainties in classification tasks.
method Combines path integrals, counterfactual explanations, and generative models.
result Framework outperforms existing alternatives in quantitative evaluations.
This paper extends explainability methods to non-Gaussian Gaussian Processes.
problem Making non-Gaussian GP models transparent and explainable.
method Proposes Integrated Gradient-based explainability for non-Gaussian GP models.
result Offers both analytical and approximate solutions for non-Gaussian GP models.
New findings show local attributions can't be both robust and provide recourse.
problem Ensuring machine learning systems are accountable and provide actionable recourse options.
method Formal definition of recourse sensitivity and counterexamples for popular attribution methods.
result It is impossible for any single attribution method to be both robust and provide recourse.
Captum library simplifies model interpretability for PyTorch.
problem Improving model interpretability for various PyTorch models.
method Unified, open-source library with generic attribution algorithms and metrics.
result Unified and scalable model interpretability for multimodal inputs.