Non-identifiability of Explanations from Model Behavior in Deep Networks of Image Authenticity Judgments

• URL: http://arxiv.org/abs/2604.07254v1
• Date: Wed, 08 Apr 2026 16:15:08 GMT
• Title: Non-identifiability of Explanations from Model Behavior in Deep Networks of Image Authenticity Judgments
• Abstract summary: Deep neural networks can predict human judgments, but this does not imply that they rely on human-like information or reveal cues underlying those judgments.<n>We tested whether models that predict human authenticity ratings also produce consistent explanations within and across architectures.<n>We conclude that while deep networks can predict human authenticity judgments well, they do not produce identifiable explanations for those judgments.
• Score: 0.07161783472741748
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Deep neural networks can predict human judgments, but this does not imply that they rely on human-like information or reveal the cues underlying those judgments. Prior work has addressed this issue using attribution heatmaps, but their explanatory value in itself depends on robustness. Here we tested the robustness of such explanations by evaluating whether models that predict human authenticity ratings also produce consistent explanations within and across architectures. We fit lightweight regression heads to multiple frozen pretrained vision models and generated attribution maps using Grad-CAM, LIME, and multiscale pixel masking. Several architectures predicted ratings well, reaching about 80% of the noise ceiling. VGG models achieved this by tracking image quality rather than authenticity-specific variance, limiting the relevance of their attributions. Among the remaining models, attribution maps were generally stable across random seeds within an architecture, especially for EfficientNetB3 and Barlow Twins, and consistency was higher for images judged as more authentic. Crucially, agreement in attribution across architectures was weak even when predictive performance was similar. To address this, we combined models in ensembles, which improved prediction of human authenticity judgments and enabled image-level attribution via pixel masking. We conclude that while deep networks can predict human authenticity judgments well, they do not produce identifiable explanations for those judgments. More broadly, our findings suggest that post hoc explanations from successful models of behavior should be treated as weak evidence for cognitive mechanism.

Related papers

• A Meaningful Perturbation Metric for Evaluating Explainability Methods [55.09730499143998]
  We introduce a novel approach, which harnesses image generation models to perform targeted perturbation.<n> Specifically, we focus on inpainting only the high-relevance pixels of an input image to modify the model's predictions while preserving image fidelity.<n>This is in contrast to existing approaches, which often produce out-of-distribution modifications, leading to unreliable results.
  arXiv  Detail & Related papers  (2025-04-09T11:46:41Z)
• Improving Network Interpretability via Explanation Consistency Evaluation [56.14036428778861]
  We propose a framework that acquires more explainable activation heatmaps and simultaneously increase the model performance. Specifically, our framework introduces a new metric, i.e., explanation consistency, to reweight the training samples adaptively in model learning. Our framework then promotes the model learning by paying closer attention to those training samples with a high difference in explanations.
  arXiv  Detail & Related papers  (2024-08-08T17:20:08Z)
• Counterfactual Image Generation for adversarially robust and interpretable Classifiers [1.3859669037499769]
  We propose a unified framework leveraging image-to-image translation Generative Adrial Networks (GANs) to produce counterfactual samples. This is achieved by combining the classifier and discriminator into a single model that attributes real images to their respective classes and flags generated images as "fake" We show how the model exhibits improved robustness to adversarial attacks, and we show how the discriminator's "fakeness" value serves as an uncertainty measure of the predictions.
  arXiv  Detail & Related papers  (2023-10-01T18:50:29Z)
• Uncertainty in AI: Evaluating Deep Neural Networks on Out-of-Distribution Images [0.0]
  This paper investigates the uncertainty of various deep neural networks, including ResNet-50, VGG16, DenseNet121, AlexNet, and GoogleNet, when dealing with perturbed data. While ResNet-50 was the most accurate single model for OOD images, the ensemble performed even better, correctly classifying all images.
  arXiv  Detail & Related papers  (2023-09-04T22:46:59Z)
• Masked Images Are Counterfactual Samples for Robust Fine-tuning [77.82348472169335]
  Fine-tuning deep learning models can lead to a trade-off between in-distribution (ID) performance and out-of-distribution (OOD) robustness. We propose a novel fine-tuning method, which uses masked images as counterfactual samples that help improve the robustness of the fine-tuning model.
  arXiv  Detail & Related papers  (2023-03-06T11:51:28Z)
• Who Explains the Explanation? Quantitatively Assessing Feature Attribution Methods [0.0]
  We propose a novel evaluation metric -- the Focus -- designed to quantify the faithfulness of explanations. We show the robustness of the metric through randomization experiments, and then use Focus to evaluate and compare three popular explainability techniques. Our results find LRP and GradCAM to be consistent and reliable, while the latter remains most competitive even when applied to poorly performing models.
  arXiv  Detail & Related papers  (2021-09-28T07:10:24Z)
• Scene Uncertainty and the Wellington Posterior of Deterministic Image Classifiers [68.9065881270224]
  We introduce the Wellington Posterior, which is the distribution of outcomes that would have been obtained in response to data that could have been generated by the same scene. We explore the use of data augmentation, dropout, ensembling, single-view reconstruction, and model linearization to compute a Wellington Posterior. Additional methods include the use of conditional generative models such as generative adversarial networks, neural radiance fields, and conditional prior networks.
  arXiv  Detail & Related papers  (2021-06-25T20:10:00Z)
• Modeling Object Dissimilarity for Deep Saliency Prediction [86.14710352178967]
  We introduce a detection-guided saliency prediction network that explicitly models the differences between multiple objects. Our approach is general, allowing us to fuse our object dissimilarities with features extracted by any deep saliency prediction network.
  arXiv  Detail & Related papers  (2021-04-08T16:10:37Z)
• Contemplating real-world object classification [53.10151901863263]
  We reanalyze the ObjectNet dataset recently proposed by Barbu et al. containing objects in daily life situations. We find that applying deep models to the isolated objects, rather than the entire scene as is done in the original paper, results in around 20-30% performance improvement.
  arXiv  Detail & Related papers  (2021-03-08T23:29:59Z)
• Overinterpretation reveals image classification model pathologies [15.950659318117694]
  convolutional neural networks (CNNs) on popular benchmarks exhibit troubling pathologies that allow them to display high accuracy even in the absence of semantically salient features. We demonstrate that neural networks trained on CIFAR-10 and ImageNet suffer from overinterpretation. Although these patterns portend potential model fragility in real-world deployment, they are in fact valid statistical patterns of the benchmark that alone suffice to attain high test accuracy.
  arXiv  Detail & Related papers  (2020-03-19T17:12:23Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.