# Can Out-of-Distribution Evaluations Uncover Reliance on Shortcuts? A Case Study in Question Answering

arXiv: https://arxiv.org/abs/2508.18407 · HTML: https://mlsanity.com/paper/2508.18407/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 20 of its 27 references that are in the map.
- First posted on arXiv 25 Aug 2025.

## Builds on (most cited first, 20 of 20)

- [Measuring Massive Multitask Language Understanding](https://mlsanity.com/paper/2009.03300/) (2020)
- [SQuAD: 100,000+ Questions for Machine Comprehension of Text](https://mlsanity.com/paper/1606.05250/) (2016)
- [TriviaQA: A Large Scale Distantly Supervised Challenge Dataset for Reading Comprehension](https://mlsanity.com/paper/1705.03551/) (2017)
- [Right for the Wrong Reasons: Diagnosing Syntactic Heuristics in Natural Language Inference](https://mlsanity.com/paper/1902.01007/) (2019)
- [Adversarial NLI: A New Benchmark for Natural Language Understanding](https://mlsanity.com/paper/1910.14599/) (2019)
- [Adversarial Examples for Evaluating Reading Comprehension Systems](https://mlsanity.com/paper/1707.07328/) (2017)
- [Larger language models do in-context learning differently](https://mlsanity.com/paper/2303.03846/) (2023)
- [NewsQA: A Machine Comprehension Dataset](https://mlsanity.com/paper/1611.09830/) (2016)
- [Don't Take the Easy Way Out: Ensemble Based Methods for Avoiding Known Dataset Biases](https://mlsanity.com/paper/1909.03683/) (2019)
- [SearchQA: A New Q&A Dataset Augmented with Context from a Search Engine](https://mlsanity.com/paper/1704.05179/) (2017)
- [Beat the AI: Investigating Adversarial Human Annotation for Reading Comprehension](https://mlsanity.com/paper/2002.00293/) (2020)
- [Measure and Improve Robustness in NLP Models: A Survey](https://mlsanity.com/paper/2112.08313/) (2021)
- [Mind the Trade-off: Debiasing NLU Models without Degrading the In-distribution Performance](https://mlsanity.com/paper/2005.00315/) (2020)
- [Learning and Evaluating General Linguistic Intelligence](https://mlsanity.com/paper/1901.11373/) (2019)
- [Towards Interpreting and Mitigating Shortcut Learning Behavior of NLU Models](https://mlsanity.com/paper/2103.06922/) (2021)
- [Examining and Combating Spurious Features under Distribution Shift](https://mlsanity.com/paper/2106.07171/) (2021)
- [Causal de Finetti: On the Identification of Invariant Causal Structure in Exchangeable Data](https://mlsanity.com/paper/2203.15756/) (2022)
- [Exploring The Landscape of Distributional Robustness for Question Answering Models](https://mlsanity.com/paper/2210.12517/) (2022)
- [Can Question Generation Debias Question Answering Models? A Case Study on Question-Context Lexical Overlap](https://mlsanity.com/paper/2109.11256/) (2021)
- [Think Twice: Measuring the Efficiency of Eliminating Prediction Shortcuts of Question Answering Models](https://mlsanity.com/paper/2305.06841/) (2023)

## Abstract (arXiv)

A majority of recent work in AI assesses models' generalization capabilities through the lens of performance on out-of-distribution (OOD) datasets. Despite their practicality, such evaluations build upon a strong assumption: that OOD evaluations can capture and reflect upon possible failures in a real-world deployment. In this work, we challenge this assumption and confront the results obtained from OOD evaluations with a set of specific failure modes documented in existing question-answering (QA) models, referred to as a reliance on spurious features or prediction shortcuts. We find that different datasets used for OOD evaluations in QA provide an estimate of models' robustness to shortcuts that have a vastly different quality, some largely under-performing even a simple, in-distribution evaluation. We partially attribute this to the observation that spurious shortcuts are shared across ID+OOD datasets, but also find cases where a dataset's quality for training and evaluation is largely disconnected. Our work underlines limitations of commonly-used OOD-based evaluations of generalization, and provides methodology and recommendations for evaluating generalization within and beyond QA more robustly.