Bias Leaves a Gradient Trail: Label-Free Bias Identification via Gradient Probes on Concept Decompositions 事件

Bias Leaves a Gradient Trail: Label-Free Bias Identification via Gradient Probes on Concept Decompositions arXiv:2605.28780v1 Announce Type: new Abstract: Vision classifiers can exploit spurious correlations, achieving high in-distribution accuracy yet failing under distribution shift. Existing approaches to bias mitigation and analysis often depend on curated datasets, spurious-attribute or group labels, or retraining, which may be infeasible once a model is deployed or the relevant bias is un