Spectral Integrated Gradients for Coarse-to-Fine Feature Attribution
For practitioners needing reliable feature attributions in image classification, SIG offers a principled coarse-to-fine path that reduces noise and enhances interpretability.
Spectral Integrated Gradients (SIG) improves feature attribution by constructing integration paths via SVD, activating global structure before fine details, resulting in cleaner attribution maps with reduced noise and improved quantitative performance over existing path-based methods.
Integrated Gradients (IG) is a widely adopted feature attribution method that satisfies desirable axiomatic properties. However, the choice of integration path significantly affects the quality of attributions, and the standard straight-line path introduces all input features simultaneously, often accumulating noisy gradients along the way. To address this limitation, we propose Spectral Integrated Gradients, which constructs integration paths based on singular value decomposition (SVD) of the baseline-to-input difference. By progressively activating singular components from largest to smallest, SIG introduces global structure before fine-grained details, naturally following a coarse-to-fine progression. Through extensive evaluation across diverse image classification datasets, we demonstrate that SIG produces cleaner attribution maps with reduced noise and achieves improved quantitative performance compared to existing path-based attribution methods. Our code is available at https://github.com/leekwoon/sig/.