无成本认知对齐：诱导人类注意力偏差以提升可解释性视觉Transformer
Cognitive Alignment At No Cost: Inducing Human Attention Biases For Interpretable Vision Transformers
AI总结：
通过微调Google ViT-B/16的自注意力权重，利用人类显著性固定图提升模型与人类注意力的对齐，发现无需额外成本即可增强模型可解释性。
摘要：
对于最先进的图像理解，视觉Transformer（ViT）已成为标准架构，但其处理方式与人类注意力特征存在显著差异。我们探讨通过微调Google ViT-B/16的自注意力权重，利用人类显著性固定图缩小这种认知差距。为区分语义相关信号与通用监督影响，微调模型与洗牌对照组进行比较。微调显著提升了五种显著性指标的对齐，并诱导出三种标志性的人类偏见：微调逆转了基线对大物体的反人类偏见，放大了动性偏好并降低了极端注意力熵。贝叶斯平衡分析提供了决定性至非常强的证据，表明这种认知对齐不会影响模型在ImageNet、ImageNet-C和Out-of-distribution（ObjectNet）基准上的原始分类性能。等效过程应用于ResNet-50卷积神经网络（CNN）时，导致对齐和准确率均下降，表明ViT的模块化自注意力机制独特地适用于区分空间优先级与表征逻辑。这些发现表明，生物基础的先验知识可以作为人类对齐注意力的自由涌现属性，以提高Transformer的可解释性。
英文摘要：
For state-of-the-art image understanding, Vision Transformers (ViTs) have become the standard architecture but their processing diverges substantially from human attentional characteristics. We investigate whether this cognitive gap can be shrunk by fine-tuning the self-attention weights of Google's ViT-B/16 on human saliency fixation maps. To isolate the effects of semantically relevant signals from generic human supervision, the tuned model is compared against a shuffled control. Fine-tuning significantly improved alignment across five saliency metrics and induced three hallmark human-like biases: tuning reversed the baseline's anti-human large-object bias toward small-objects, amplified the animacy preference and diminished extreme attention entropy. Bayesian parity analysis provides decisive to very-strong evidence that this cognitive alignment comes at no cost to the model's original classification performance on in- (ImageNet), corrupted (ImageNet-C) and out-of-distribution (ObjectNet) benchmarks. An equivalent procedure applied to a ResNet-50 Convolutional Neural Network (CNN) instead degraded both alignment and accuracy, suggesting that the ViT's modular self-attention mechanism is uniquely suited for dissociating spatial priority from representational logic. These findings demonstrate that biologically grounded priors can be instilled as a free emergent property of human-aligned attention, to improve transformer interpretability.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
