Abstract
Worst-group accuracy (WGA) evaluates a trained predictor but does not characterize how its frozen backbone behaves when a new head is learned. We introduce BiasFlow, a hook-based toolkit for monitoring class-attribute centroid alignment (IBMI), within-class centroid separation (W-IBMI), and feature-projection sensitivity. IBMI is confounded by class-attribute correlation and is not a measure of causal feature reliance. We pair these diagnostics with BiasFlow Regularization (BFR), a supervised, composable class-conditional centroid-alignment penalty. W-IBMI verifies the quantity BFR optimizes; it is scale dependent and does not independently establish attribute removal. Across the reported small-scale benchmarks, adding BFR improves or preserves mean WGA, with gains up to +26.0 pp on UrbanCars. The principal independent stress test freezes CelebA-Std backbones and trains fresh heads on biased data: BFR+GroupDRO improves WGA from 40.7% to 64.1%, while Male probe accuracy decreases from 92.5% to 72.2%. Attribute information remains recoverable, and cross-task results are mixed. A controlled synthetic-watermark ImageNet experiment additionally improves watermark-shift accuracy by +23.0 pp under matched training. These results support evaluating centroid geometry and resistance to biased head retraining alongside WGA, within the tested protocols.
Keywords
Subject
Publication details
- Journal
- Not available
- Open access
- Green open access
Cite this article
APA 7
Deng, H., Lin, Z., & Yang, Y. (2026). BiasFlow: Geometric Monitoring and Backbone Regularization for Spurious Feature Reliance. https://omanscience.com/en/articles/biasflow-geometric-monitoring-and-backbone-regularization-for-spurious-feature-reliance
MLA 9
Deng, Haojin, et al. "BiasFlow: Geometric Monitoring and Backbone Regularization for Spurious Feature Reliance." https://omanscience.com/en/articles/biasflow-geometric-monitoring-and-backbone-regularization-for-spurious-feature-reliance.
Chicago (author–date)
Deng, Haojin, Zhiping Lin, and Yimin Yang. 2026. "BiasFlow: Geometric Monitoring and Backbone Regularization for Spurious Feature Reliance." https://omanscience.com/en/articles/biasflow-geometric-monitoring-and-backbone-regularization-for-spurious-feature-reliance.
Harvard
Deng, H., Lin, Z. and Yang, Y. (2026) 'BiasFlow: Geometric Monitoring and Backbone Regularization for Spurious Feature Reliance', Available at: https://omanscience.com/en/articles/biasflow-geometric-monitoring-and-backbone-regularization-for-spurious-feature-reliance.
Vancouver
Deng H, Lin Z, Yang Y. BiasFlow: Geometric Monitoring and Backbone Regularization for Spurious Feature Reliance. https://omanscience.com/en/articles/biasflow-geometric-monitoring-and-backbone-regularization-for-spurious-feature-reliance
IEEE
H. Deng, Z. Lin, and Y. Yang, "BiasFlow: Geometric Monitoring and Backbone Regularization for Spurious Feature Reliance," https://omanscience.com/en/articles/biasflow-geometric-monitoring-and-backbone-regularization-for-spurious-feature-reliance.