Counterfactual explanations and algorithmic recourses for machine learning: A review
Machine learning plays a role in many deployed decision systems, often in ways that are
difficult or impossible to understand by human stakeholders. Explaining, in a human …
difficult or impossible to understand by human stakeholders. Explaining, in a human …
Mechanistic mode connectivity
We study neural network loss landscapes through the lens of mode connectivity, the
observation that minimizers of neural networks retrieved via training on a dataset are …
observation that minimizers of neural networks retrieved via training on a dataset are …
Diffusion models for counterfactual explanations
Counterfactual explanations have shown promising results as a post-hoc framework to make
image classifiers more explainable. In this paper, we propose DiME, a method allowing the …
image classifiers more explainable. In this paper, we propose DiME, a method allowing the …
Interpretability for reliable, efficient, and self-cognitive DNNs: From theories to applications
In recent years, remarkable achievements have been made in artificial intelligence tasks
and applications based on deep neural networks (DNNs), especially in the fields of vision …
and applications based on deep neural networks (DNNs), especially in the fields of vision …
Adversarial counterfactual visual explanations
Counterfactual explanations and adversarial attacks have a related goal: flipping output
labels with minimal perturbations regardless of their characteristics. Yet, adversarial attacks …
labels with minimal perturbations regardless of their characteristics. Yet, adversarial attacks …
Predicting is not understanding: Recognizing and addressing underspecification in machine learning
Abstract Machine learning (ML) models are typically optimized for their accuracy on a given
dataset. However, this predictive criterion rarely captures all desirable properties of a model …
dataset. However, this predictive criterion rarely captures all desirable properties of a model …
Quantum machine learning for audio classification with applications to healthcare
Accessible rapid COVID-19 testing continues to be necessary and several studies involving
deep neural network (DNN) methods for detection have been published. As part of a …
deep neural network (DNN) methods for detection have been published. As part of a …
A generalized explanation framework for visualization of deep learning model predictions
P Wang, N Vasconcelos - IEEE Transactions on Pattern …, 2023 - ieeexplore.ieee.org
Attribution-based explanations are popular in computer vision but of limited use for fine-
grained classification problems typical of expert domains, where classes differ by subtle …
grained classification problems typical of expert domains, where classes differ by subtle …
Latent Diffusion Counterfactual Explanations
Counterfactual explanations have emerged as a promising method for elucidating the
behavior of opaque black-box models. Recently, several works leveraged pixel-space …
behavior of opaque black-box models. Recently, several works leveraged pixel-space …
MAGANet: Achieving combinatorial generalization by modeling a group action
Combinatorial generalization refers to the ability to collect and assemble various attributes
from diverse data to generate novel unexperienced data. This ability is considered a …
from diverse data to generate novel unexperienced data. This ability is considered a …