Invariant Risk Minimization
method
ID: invariantriskminimization
Related Papers
Improved Multi-label Classification under Temporal Concept Drift: Rethinking Group-Robust Algorithms in a Label-Wise Setting
AbstractIn document classification for, e.g., legal and biomedical text, we often deal with hundreds of classes, including very infrequent ones, as well as temporal concept drift caused by the influence …...
Causal-Debias: Unifying Debiasing in Pretrained Language Models and Fine-tuning via Causal Invariant Learning
AbstractDemographic biases and social stereotypes are common in pretrained language models (PLMs), and a burgeoning body of literature focuses on removing the unwanted stereotypical associations from PLMs. However, when fine-tuning …...