# Rethinking Feature Reliance Evaluation with Semantically Matched Suppression

Source: [Rethinking Feature Reliance Evaluation with Semantically Matched Suppression](<https://www.sota2.com/research/paper/6a5efd0a17e2c828503b9f78>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-07-22T05:00:05.108Z

Authors: Ning Jiang, Tianyi Luo, Zhengyong Huang, Yao Sui (1 and 2 and 3) \_\_INSTITUTION\_4\_\_ Institute of Medical Technology, Peking University Health Science Center, Beijing, China, (2) National Institute of Health Data Science, Peking University, Beijing, China, (3) Institute for Artificial Intelligence, Peking University, Beijing, China, (4) School of Computer Science, Engineering, Sun Yat-sen University, Guangzhou, China)

Published: 2026-07-13T13:02:07.000Z

## Abstract

Understanding whether visual recognition models rely on shape, texture, or color is central to interpreting their behavior. Prior cue-conflict studies have strongly influenced the view that CNNs are texture-biased, yet such tests measure cue preference under artificial conflicts rather than feature reliance during natural recognition. We revisit this question through controlled feature suppression and show that performance drops are difficult to interpret unless different suppression operations impose comparable category-level damage. We introduce a semantically matched evaluation framework that compares shape and texture suppression at matched levels of category separability loss. Under this framework, ImageNet-trained CNNs show stronger degradation under texture suppression than under shape suppression, revealing greater texture reliance than suggested by unmatched suppression analyses. Extending the comparison across architectures, we find that Vision Transformers retain higher accuracy than CNNs under both shape and texture suppression. Brain encoding further shows that ViT representations exhibit smaller suppression-induced decreases in neural prediction performance under the tested suppression settings. These findings indicate that semantic comparability is essential for interpreting feature reliance from suppression experiments, and suggest that the robustness advantage of ViTs may be related to representations more compatible with human visual cortex.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2607.16298>)

[PDF](<https://arxiv.org/pdf/2607.16298.pdf>)
