Activation Matching for Explanation Generation

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Suhail, Pirzada, Anand, Aditya, Sethi, Amit
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866908617930702848
author Suhail, Pirzada
Anand, Aditya
Sethi, Amit
author_facet Suhail, Pirzada
Anand, Aditya
Sethi, Amit
contents In this paper we introduce an activation-matching--based approach to generate minimal, faithful explanations for the decision-making of a pretrained classifier on any given image. Given an input image $x$ and a frozen model $f$, we train a lightweight autoencoder to output a binary mask $m$ such that the explanation $e = m \odot x$ preserves both the model's prediction and the intermediate activations of \(x\). Our objective combines: (i) multi-layer activation matching with KL divergence to align distributions and cross-entropy to retain the top-1 label for both the image and the explanation; (ii) mask priors -- L1 area for minimality, a binarization penalty for crisp 0/1 masks, and total variation for compactness; and (iii) abductive constraints for faithfulness and necessity. Together, these objectives yield small, human-interpretable masks that retain classifier behavior while discarding irrelevant input regions, providing practical and faithful minimalist explanations for the decision making of the underlying model.
format Preprint
id arxiv_https___arxiv_org_abs_2509_23051
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Activation Matching for Explanation Generation
Suhail, Pirzada
Anand, Aditya
Sethi, Amit
Computer Vision and Pattern Recognition
Machine Learning
In this paper we introduce an activation-matching--based approach to generate minimal, faithful explanations for the decision-making of a pretrained classifier on any given image. Given an input image $x$ and a frozen model $f$, we train a lightweight autoencoder to output a binary mask $m$ such that the explanation $e = m \odot x$ preserves both the model's prediction and the intermediate activations of \(x\). Our objective combines: (i) multi-layer activation matching with KL divergence to align distributions and cross-entropy to retain the top-1 label for both the image and the explanation; (ii) mask priors -- L1 area for minimality, a binarization penalty for crisp 0/1 masks, and total variation for compactness; and (iii) abductive constraints for faithfulness and necessity. Together, these objectives yield small, human-interpretable masks that retain classifier behavior while discarding irrelevant input regions, providing practical and faithful minimalist explanations for the decision making of the underlying model.
title Activation Matching for Explanation Generation
topic Computer Vision and Pattern Recognition
Machine Learning
url https://arxiv.org/abs/2509.23051