On the Reliability of Cue Conflict and Beyond

TL;DR

This paper introduces REFINED-BIAS, a new dataset and framework for reliably diagnosing shape and texture biases in neural networks, addressing limitations of previous stylization-based methods and enabling fairer, more interpretable comparisons.

Contribution

The paper presents REFINED-BIAS, an improved evaluation framework that constructs balanced cue pairs and measures cue sensitivity across the full label space for better bias diagnosis.

Findings

REFINED-BIAS provides more consistent bias estimates.

It enables fairer cross-model comparisons.

It clarifies empirical conclusions about shape and texture biases.

Abstract

Understanding how neural networks rely on visual cues offers a human-interpretable view of their internal decision processes. The cue-conflict benchmark has been influential in probing shape-texture preference and in motivating the insight that stronger, human-like shape bias is often associated with improved in-domain performance. However, we find that the current stylization-based instantiation can yield unstable and ambiguous bias estimates. Specifically, stylization may not reliably instantiate perceptually valid and separable cues nor control their relative informativeness, ratio-based bias can obscure absolute cue sensitivity, and restricting evaluation to preselected classes can distort model predictions by ignoring the full decision space. Together, these factors can confound preference with cue validity, cue balance, and recognizability artifacts. We introduce REFINED-BIAS, an integrated dataset and evaluation framework for reliable and interpretable shape-texture bias diagnosis. REFINED-BIAS constructs balanced, human- and model- recognizable cue pairs using explicit definitions of shape and texture, and measures cue-specific sensitivity over the full label space via a ranking-based metric, enabling fairer cross-model comparisons. Across diverse training regimes and architectures, REFINED-BIAS enables fairer cross-model comparison, more faithful diagnosis of shape and texture biases, and clearer empirical conclusions, resolving inconsistencies that prior cue-conflict evaluations could not reliably disambiguate.