# Generating Data to Mitigate Spurious Correlations in Natural Language Inference Datasets

arXiv: https://arxiv.org/abs/2203.12942 · HTML: https://mlsanity.com/paper/2203.12942/

## Key facts

- Cited by 54 papers in the map (data through 14 Oct 2026).
- Builds on 41 of its 52 references that are in the map.
- First posted on arXiv 24 Mar 2022.

## Builds on (most cited first, 30 of 41)

- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [RoBERTa: A Robustly Optimized BERT Pretraining Approach](https://mlsanity.com/paper/1907.11692/) (2019)
- [ALBERT: A Lite BERT for Self-supervised Learning of Language Representations](https://mlsanity.com/paper/1909.11942/) (2019)
- [A Broad-Coverage Challenge Corpus for Sentence Understanding through Inference](https://mlsanity.com/paper/1704.05426/) (2017)
- [A large annotated corpus for learning natural language inference](https://mlsanity.com/paper/1508.05326/) (2015)
- [FEVER: a large-scale dataset for Fact Extraction and VERification](https://mlsanity.com/paper/1803.05355/) (2018)
- [Right for the Wrong Reasons: Diagnosing Syntactic Heuristics in Natural Language Inference](https://mlsanity.com/paper/1902.01007/) (2019)
- [Annotation Artifacts in Natural Language Inference Data](https://mlsanity.com/paper/1803.02324/) (2018)
- [Neural Text Generation with Unlikelihood Training](https://mlsanity.com/paper/1908.04319/) (2019)
- [Hypothesis Only Baselines in Natural Language Inference](https://mlsanity.com/paper/1805.01042/) (2018)
- [Don't Take the Easy Way Out: Ensemble Based Methods for Avoiding Known Dataset Biases](https://mlsanity.com/paper/1909.03683/) (2019)
- [Stress Test Evaluation for Natural Language Inference](https://mlsanity.com/paper/1806.00692/) (2018)
- [PAQ: 65 Million Probably-Asked Questions and What You Can Do With Them](https://mlsanity.com/paper/2102.07033/) (2021)
- [Generating Datasets with Pretrained Language Models](https://mlsanity.com/paper/2104.07540/) (2021)
- [Are We Modeling the Task or the Annotator? An Investigation of Annotator Bias in Natural Language Understanding Datasets](https://mlsanity.com/paper/1908.07898/) (2019)
- [Adversarial Filters of Dataset Biases](https://mlsanity.com/paper/2002.04108/) (2020)
- [Towards Debiasing Fact Verification Models](https://mlsanity.com/paper/1908.05267/) (2019)
- [Breaking NLI Systems with Sentences that Require Simple Lexical Inferences](https://mlsanity.com/paper/1805.02266/) (2018)
- [End-to-End Bias Mitigation by Modelling Biases in Corpora](https://mlsanity.com/paper/1909.06321/) (2019)
- [Unlearn Dataset Bias in Natural Language Inference by Fitting the Residual](https://mlsanity.com/paper/1908.10763/) (2019)
- [Competency Problems: On Finding and Removing Artifacts in Language Data](https://mlsanity.com/paper/2104.08646/) (2021)
- [Mind the Trade-off: Debiasing NLU Models without Degrading the In-distribution Performance](https://mlsanity.com/paper/2005.00315/) (2020)
- [Learning from others' mistakes: Avoiding dataset biases without modeling them](https://mlsanity.com/paper/2012.01300/) (2020)
- [Improving Question Answering Model Robustness with Synthetic Adversarial Data Generation](https://mlsanity.com/paper/2104.08678/) (2021)
- [A Thorough Examination of the CNN/Daily Mail Reading Comprehension Task](https://mlsanity.com/paper/1606.02858/) (2016)
- [Tailor: Generating and Perturbing Text with Semantic Controls](https://mlsanity.com/paper/2107.07150/) (2021)
- [Generalization in NLI: Ways (Not) To Go Beyond Simple Heuristics](https://mlsanity.com/paper/2110.01518/) (2021)
- [Don't Take the Premise for Granted: Mitigating Artifacts in Natural Language Inference](https://mlsanity.com/paper/1907.04380/) (2019)
- [Generative Data Augmentation for Commonsense Reasoning](https://mlsanity.com/paper/2004.11546/) (2020)
- [Adversarially Regularising Neural NLI Models to Integrate Logical Background Knowledge](https://mlsanity.com/paper/1808.08609/) (2018)

## Built on by (most cited first, 30 of 54)

- [Challenges and Applications of Large Language Models](https://mlsanity.com/paper/2307.10169/) (2023)
- [Are We Done with MMLU?](https://mlsanity.com/paper/2406.04127/) (2024)
- [Shortcut Learning of Large Language Models in Natural Language Understanding](https://mlsanity.com/paper/2208.11857/) (2022)
- [Natural Language Reasoning, A Survey](https://mlsanity.com/paper/2303.14725/) (2023)
- [BITE: Textual Backdoor Attacks with Iterative Trigger Injection](https://mlsanity.com/paper/2205.12700/) (2022)
- [Language Models in the Loop: Incorporating Prompting into Weak Supervision](https://mlsanity.com/paper/2205.02318/) (2022)
- [Privacy in Large Language Models: Attacks, Defenses and Future Directions](https://mlsanity.com/paper/2310.10383/) (2023)
- [Stretching Sentence-pair NLI Models to Reason over Long Documents and Clusters](https://mlsanity.com/paper/2204.07447/) (2022)
- [InjecGuard: Benchmarking and Mitigating Over-defense in Prompt Injection Guardrail Models](https://mlsanity.com/paper/2410.22770/) (2024)
- [Neuro-Symbolic AI in 2024: A Systematic Review](https://mlsanity.com/paper/2501.05435/) (2025)
- [DISCO: Distilling Counterfactuals with Large Language Models](https://mlsanity.com/paper/2212.10534/) (2022)
- [Towards Trustworthy and Aligned Machine Learning: A Data-centric Survey with Causality Perspectives](https://mlsanity.com/paper/2307.16851/) (2023)
- [Explanation-based Finetuning Makes Models More Robust to Spurious Cues](https://mlsanity.com/paper/2305.04990/) (2023)
- [CORE: A Retrieve-then-Edit Framework for Counterfactual Data Generation](https://mlsanity.com/paper/2210.04873/) (2022)
- [Mitigating Backdoor Poisoning Attacks through the Lens of Spurious Correlation](https://mlsanity.com/paper/2305.11596/) (2023)
- [Extracting or Guessing? Improving Faithfulness of Event Temporal Relation Extraction](https://mlsanity.com/paper/2210.04992/) (2022)
- [CEval: A Benchmark for Evaluating Counterfactual Text Generation](https://mlsanity.com/paper/2404.17475/) (2024)
- [Can NLI Provide Proper Indirect Supervision for Low-resource Biomedical Relation Extraction?](https://mlsanity.com/paper/2212.10784/) (2022)
- [Improving Conversational Recommendation Systems via Bias Analysis and Language-Model-Enhanced Data Augmentation](https://mlsanity.com/paper/2310.16738/) (2023)
- [AGRO: Adversarial Discovery of Error-prone groups for Robust Optimization](https://mlsanity.com/paper/2212.00921/) (2022)
- [Feature-Level Debiased Natural Language Understanding](https://mlsanity.com/paper/2212.05421/) (2022)
- [Does Self-Rationalization Improve Robustness to Spurious Correlations?](https://mlsanity.com/paper/2210.13575/) (2022)
- [Alfred: A System for Prompted Weak Supervision](https://mlsanity.com/paper/2305.18623/) (2023)
- [ABEX: Data Augmentation for Low-Resource NLU via Expanding Abstract Descriptions](https://mlsanity.com/paper/2406.04286/) (2024)
- [FactMix: Using a Few Labeled In-domain Examples to Generalize to Cross-domain Named Entity Recognition](https://mlsanity.com/paper/2208.11464/) (2022)
- [Hierarchical Indexing for Retrieval-Augmented Opinion Summarization](https://mlsanity.com/paper/2403.00435/) (2024)
- [LUCID: LLM-Generated Utterances for Complex and Interesting Dialogues](https://mlsanity.com/paper/2403.00462/) (2024)
- [AutoCAD: Automatically Generating Counterfactuals for Mitigating Shortcut Learning](https://mlsanity.com/paper/2211.16202/) (2022)
- [RFiD: Towards Rational Fusion-in-Decoder for Open-Domain Question Answering](https://mlsanity.com/paper/2305.17041/) (2023)
- [Learning Shortcuts: On the Misleading Promise of NLU in Language Models](https://mlsanity.com/paper/2401.09615/) (2024)

## Abstract (arXiv)

Natural language processing models often exploit spurious correlations between task-independent features and labels in datasets to perform well only within the distributions they are trained on, while not generalising to different task distributions. We propose to tackle this problem by generating a debiased version of a dataset, which can then be used to train a debiased, off-the-shelf model, by simply replacing its training data. Our approach consists of 1) a method for training data generators to generate high-quality, label-consistent data samples; and 2) a filtering mechanism for removing data points that contribute to spurious correlations, measured in terms of z-statistics. We generate debiased versions of the SNLI and MNLI datasets, and we evaluate on a large suite of debiased, out-of-distribution, and adversarial test sets. Results show that models trained on our debiased datasets generalise better than those trained on the original datasets in all settings. On the majority of the datasets, our method outperforms or performs comparably to previous state-of-the-art debiasing strategies, and when combined with an orthogonal technique, product-of-experts, it improves further and outperforms previous best results of SNLI-hard and MNLI-hard.