AutoCAD: Automatically Generating Counterfactuals for Mitigating Shortcut Learning
This addresses the issue of shortcut learning in NLU models for researchers and practitioners, offering a fully automatic and task-agnostic solution that is incremental over existing CAD methods.
The paper tackles the problem of reducing NLU models' reliance on spurious features by automatically generating counterfactually augmented data, achieving comparable or better out-of-distribution performance than previous human-in-the-loop or task-specific methods across multiple benchmarks.
Recent studies have shown the impressive efficacy of counterfactually augmented data (CAD) for reducing NLU models' reliance on spurious features and improving their generalizability. However, current methods still heavily rely on human efforts or task-specific designs to generate counterfactuals, thereby impeding CAD's applicability to a broad range of NLU tasks. In this paper, we present AutoCAD, a fully automatic and task-agnostic CAD generation framework. AutoCAD first leverages a classifier to unsupervisedly identify rationales as spans to be intervened, which disentangles spurious and causal features. Then, AutoCAD performs controllable generation enhanced by unlikelihood training to produce diverse counterfactuals. Extensive evaluations on multiple out-of-domain and challenge benchmarks demonstrate that AutoCAD consistently and significantly boosts the out-of-distribution performance of powerful pre-trained models across different NLU tasks, which is comparable or even better than previous state-of-the-art human-in-the-loop or task-specific CAD methods. The code is publicly available at https://github.com/thu-coai/AutoCAD.