Search papers, labs, and topics across Lattice.
The paper identifies feature correlation in neural network representations as a key cause of noisy and unreliable saliency maps. To address this, they introduce SaliencyDecor, a training framework that enforces feature decorrelation through a novel regularization term during training, without modifying the model architecture or saliency method. Experiments show SaliencyDecor produces sharper, more object-focused saliency maps and, surprisingly, also improves predictive accuracy across multiple benchmarks.
Feature decorrelation during training not only sharpens saliency maps, but also *improves* model accuracy, challenging the conventional wisdom that interpretability comes at the cost of performance.
Gradient-based saliency methods are widely used to interpret deep neural networks, yet they often produce noisy and unstable explanations that poorly align with semantically meaningful input features. We argue that a fundamental cause of this behavior lies in the geometry of learned representations: correlated feature dimensions diffuse attribution gradients across redundant directions, resulting in blurred and unreliable saliency maps. To address this issue, we identify feature correlation as a structural limitation of gradient-based interpretability and propose SaliencyDecor, a training framework that enforces feature decorrelation to improve attribution fidelity without modifying saliency methods or model architectures by reshaping the feature space toward orthogonality, our approach promotes more concentrated gradient flow and improves the fidelity of saliency-based explanations. SaliencyDecor jointly optimizes classification, prediction consistency under feature masking, and a decorrelation regularizer, requiring no architectural changes or inference-time overhead. Extensive experiments across multiple benchmarks and architectures demonstrate that our method produces substantially sharper and more object-focused saliency maps while simultaneously improving predictive performance, achieving accuracy gains across the datasets. These results establish our method as a principled mechanism for enhancing both interpretability and accuracy, challenging the conventional trade-off between explanation quality and model performance.
