低级线索早期精度影响视觉模型捷径学习，越可靠越易依赖错误线索。
Early Cue Precision Shapes Visual Shortcut Learning in Controlled Cue-Manipulation Benchmarks
- 通过操控线索预测准确性，研究早期学习阶段线索可靠性对模型行为的影响。
- 线索越精确，冲突场景下准确率下降越明显，最高达0.589降至0.005。
- 模型在下游适应中需持续保持线索去相关，否则前期训练优势会被快速覆盖。
视觉分类器可在依赖低级线索的情况下实现高匹配分布准确率，但在冲突或抑制条件下失效。本文测试这种失效是否由早期线索精度（即低级线索在早期学习或下游探测中预测标签的可靠性）决定。在合成形状-纹理任务、序列数字训练、10类冻结表征审计以及基于CIFAR-10自然图像的纹理叠加基准中，我们操纵物体-纹理匹配概率，评估匹配身份准确率、冲突准确率、纹理选择率和抑制行为。结果表明，虽不可靠但具预测性的输入无法替代线索去相关性。在10类数字探测中，冲突准确率从随机水平的0.589降至目标完美纹理下的0.005；在CIFAR-10冻结探测中，冲突准确率从0.569降至0.114，纹理选择率从0.049升至0.855；该趋势在纹理叠加强度α∈{0.15,0.25,0.35,0.50}下均成立。端到端CIFAR-10训练显示，早期线索精度低可改善冲突前行为，但富含捷径的微调会迅速抵消此益处。因此，线索去相关必须在下游适应中持续维持，而非一次性处理。
原文摘要 · Abstract (English)
Visual classifiers can achieve high matched-distribution accuracy while relying on low-level cues that fail under conflict or suppression. We test whether this failure is shaped by early cue precision: the reliability with which a low-level cue predicts the label during early learning or downstream probe fitting. Across synthetic shape-texture tasks, sequential digit training, a 10-class frozen-representation audit, and a CIFAR-10 natural-image-based texture-overlay benchmark, we manipulate object-texture match probability and evaluate matched-ID accuracy, conflict accuracy, texture-choice rate, and suppression behavior. Degraded-but-predictive input does not substitute for cue decorrelation. In 10-class digit probes, conflict accuracy drops from 0.589 under chance-like cue precision to 0.005 under target-perfect texture. In CIFAR-10 frozen probes, conflict accuracy drops from 0.569 to 0.114, while texture choice rises from 0.049 to 0.855; this ordering persists across texture-overlay strengths alpha in {0.15,0.25,0.35,0.50}. End-to-end CIFAR-10 training shows that low early cue precision improves pre-target conflict behavior, but shortcut-rich fine-tuning can rapidly overwrite this benefit. Cue decorrelation must therefore be maintained during downstream adaptation rather than treated as a one-time inoculation.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。