REFINED-BIAS: a REliable Framework for INtegrated Evaluation and Disentangled Benchmark of Interpretable Alignment of Shape/texture in neural networks Official Code
classification evaluation-framework self-supervised-learning contrastive-learning perceptual-bias model-behavior-analysis
-
Updated
Jan 19, 2026 - Python