arXiv:2611.17107 · v1
Disentangling Spurious Correlations: How IRM Handles Diverse Shortcuts in Text Classification
摘要
In NLP, models trained via Empirical Risk Minimization (ERM) often exploit spurious correlations, such as dialectal markers, at the expense of robust linguistic features, thereby degrading their out-of-distribution (OOD) generalization. Invariant Risk Minimization (IRM) aims to suppress such shortcuts by identifying predictive relationships that remain stable across environments, but its practical utility for NLP remains debated. We use a causality-driven framework to evaluate IRM under three shortcut-generating mechanisms: label-generated features, confounding, and selection bias. Our results show that IRM successfully rejects shortcuts across the evaluated causal and anti-causal structures, including under limited environmental variation and increasing label noise. We then extend our analysis beyond controlled synthetic artifacts to two less constrained settings: gender-profession associations in Bias in Bios and semantically loaded actor-name shortcuts in ActorCorr. These experiments show that invariant learning can mitigate more realistic textual shortcuts, while semantically meaningful features require careful calibration of the invariance penalty.
作者备注：21 pages. Published at EMNLP 2026 (Main Conference)
期刊信息：Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing (EMNLP), pp. 60785-60805 (2026)