Practical Benefits of Feature Feedback Under Distribution Shift
This addresses the problem of improving model generalization for practical applications where data distributions shift, though it is incremental as it builds on prior work on counterfactual data augmentation.
The paper investigated whether feature feedback (auxiliary annotations like bounding boxes or salient text spans) improves model robustness under distribution shift, finding that it significantly enhances out-of-domain performance for sentiment analysis tasks, with comparable in-domain results, but shows no benefit for natural language inference.
In attempts to develop sample-efficient and interpretable algorithms, researcher have explored myriad mechanisms for collecting and exploiting feature feedback (or rationales) auxiliary annotations provided for training (but not test) instances that highlight salient evidence. Examples include bounding boxes around objects and salient spans in text. Despite its intuitive appeal, feature feedback has not delivered significant gains in practical problems as assessed on iid holdout sets. However, recent works on counterfactually augmented data suggest an alternative benefit of supplemental annotations, beyond interpretability: lessening sensitivity to spurious patterns and consequently delivering gains in out-of-domain evaluations. We speculate that while existing methods for incorporating feature feedback have delivered negligible in-sample performance gains, they may nevertheless provide out-of-domain benefits. Our experiments addressing sentiment analysis, show that feature feedback methods perform significantly better on various natural out-of-domain datasets despite comparable in-domain evaluations. By contrast, performance on natural language inference remains comparable. Finally, we compare those tasks where feature feedback does (and does not) help.