# Are All Spurious Features in Natural Language Alike? An Analysis through a Causal Lens

arXiv: https://arxiv.org/abs/2210.14011 · HTML: https://mlsanity.com/paper/2210.14011/

## Key facts

- Cited by 28 papers in the map (data through 14 Oct 2026).
- Builds on 30 of its 41 references that are in the map.
- First posted on arXiv 25 Oct 2022.

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

- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [Adam: A Method for Stochastic Optimization](https://mlsanity.com/paper/1412.6980/) (2014)
- [RoBERTa: A Robustly Optimized BERT Pretraining Approach](https://mlsanity.com/paper/1907.11692/) (2019)
- [GLUE: A Multi-Task Benchmark and Analysis Platform for Natural Language Understanding](https://mlsanity.com/paper/1804.07461/) (2018)
- [SQuAD: 100,000+ Questions for Machine Comprehension of Text](https://mlsanity.com/paper/1606.05250/) (2016)
- [A Broad-Coverage Challenge Corpus for Sentence Understanding through Inference](https://mlsanity.com/paper/1704.05426/) (2017)
- [HuggingFace's Transformers: State-of-the-art Natural Language Processing](https://mlsanity.com/paper/1910.03771/) (2019)
- [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)
- [Mitigating Unwanted Biases with Adversarial Learning](https://mlsanity.com/paper/1801.07593/) (2018)
- [Null It Out: Guarding Protected Attributes by Iterative Nullspace Projection](https://mlsanity.com/paper/2004.07667/) (2020)
- [Learning the Difference that Makes a Difference with Counterfactually-Augmented Data](https://mlsanity.com/paper/1909.12434/) (2019)
- [An Investigation of Why Overparameterization Exacerbates Spurious Correlations](https://mlsanity.com/paper/2005.04345/) (2020)
- [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)
- [Information-Theoretic Probing with Minimum Description Length](https://mlsanity.com/paper/2003.12298/) (2020)
- [REPAIR: Removing Representation Bias by Dataset Resampling](https://mlsanity.com/paper/1904.07911/) (2019)
- [An Empirical Study on Robustness to Spurious Correlations using Pre-trained Language Models](https://mlsanity.com/paper/2007.06778/) (2020)
- [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)
- [Desiderata for Representation Learning: A Causal Perspective](https://mlsanity.com/paper/2109.03795/) (2021)
- [Towards Unifying Feature Attribution and Counterfactual Explanations: Different Means to the Same End](https://mlsanity.com/paper/2011.04917/) (2020)
- [Identifying Spurious Correlations for Robust Text Classification](https://mlsanity.com/paper/2010.02458/) (2020)
- [Explaining Black-Box Algorithms Using Probabilistic Contrastive Counterfactuals](https://mlsanity.com/paper/2103.11972/) (2021)
- [Out-of-distribution Generalization in the Presence of Nuisance-Induced Spurious Correlations](https://mlsanity.com/paper/2107.00520/) (2021)
- [TaxiNLI: Taking a Ride up the NLU Hill](https://mlsanity.com/paper/2009.14505/) (2020)
- [On the Limitations of Dataset Balancing: The Lost Battle Against Spurious Correlations](https://mlsanity.com/paper/2204.12708/) (2022)
- [What's in a Name? Are BERT Named Entity Representations just as Good for any other Name?](https://mlsanity.com/paper/2007.06897/) (2020)
- [Debiasing Methods in Natural Language Understanding Make Bias More Accessible](https://mlsanity.com/paper/2109.04095/) (2021)

## Built on by (most cited first, 28 of 28)

- [Change is Hard: A Closer Look at Subpopulation Shift](https://mlsanity.com/paper/2302.12254/) (2023)
- [Measuring Inductive Biases of In-Context Learning with Underspecified Demonstrations](https://mlsanity.com/paper/2305.13299/) (2023)
- [Towards Trustworthy and Aligned Machine Learning: A Data-centric Survey with Causality Perspectives](https://mlsanity.com/paper/2307.16851/) (2023)
- [LLMs Are Prone to Fallacies in Causal Inference](https://mlsanity.com/paper/2406.12158/) (2024)
- [Towards Trustworthy Explanation: On Causal Rationalization](https://mlsanity.com/paper/2306.14115/) (2023)
- [Shift-Robust Molecular Relational Learning with Causal Substructure](https://mlsanity.com/paper/2305.18451/) (2023)
- [Are Human Explanations Always Helpful? Towards Objective Evaluation of Human Natural Language Explanations](https://mlsanity.com/paper/2305.03117/) (2023)
- [Understanding and Mitigating Spurious Correlations in Text Classification with Neighborhood Analysis](https://mlsanity.com/paper/2305.13654/) (2023)
- [Mind the Graph When Balancing Data for Fairness or Robustness](https://mlsanity.com/paper/2406.17433/) (2024)
- [RATE: Causal Explainability of Reward Models with Imperfect Counterfactuals](https://mlsanity.com/paper/2410.11348/) (2024)
- [Removing Spurious Concepts from Neural Network Representations via Joint Subspace Estimation](https://mlsanity.com/paper/2310.11991/) (2023)
- [Towards Practical Tool Usage for Continually Learning LLMs](https://mlsanity.com/paper/2404.09339/) (2024)
- [DECIDER: Leveraging Foundation Model Priors for Improved Model Failure Detection and Explanation](https://mlsanity.com/paper/2408.00331/) (2024)
- [Controlling Learned Effects to Reduce Spurious Correlations in Text Classifiers](https://mlsanity.com/paper/2305.16863/) (2023)
- [Debiasing Multimodal Models via Causal Information Minimization](https://mlsanity.com/paper/2311.16941/) (2023)
- [The Limits of Fair Medical Imaging AI In The Wild](https://mlsanity.com/paper/2312.10083/) (2023)
- [Looking at the Overlooked: An Analysis on the Word-Overlap Bias in Natural Language Inference](https://mlsanity.com/paper/2211.03862/) (2022)
- [Concept-Based Explanations to Test for False Causal Relationships Learned by Abusive Language Classifiers](https://mlsanity.com/paper/2307.01900/) (2023)
- [Optimizing importance weighting in the presence of sub-population shifts](https://mlsanity.com/paper/2410.14315/) (2024)
- [Zero-Shot Cross-Lingual Sentiment Classification under Distribution Shift: an Exploratory Study](https://mlsanity.com/paper/2311.06549/) (2023)
- [Discovering Highly Influential Shortcut Reasoning: An Automated Template-Free Approach](https://mlsanity.com/paper/2312.09718/) (2023)
- [CASCADE Your Datasets for Cross-Mode Knowledge Retrieval of Language Models](https://mlsanity.com/paper/2504.01450/) (2025)
- [VISPUR: Visual Aids for Identifying and Interpreting Spurious Associations in Data-Driven Decisions](https://mlsanity.com/paper/2307.14448/) (2023)
- [CoBA: Counterbias Text Augmentation for Mitigating Various Spurious Correlations via Semantic Triples](https://mlsanity.com/paper/2508.21083/) (2025)
- [De-paradox Tree: Breaking Down Simpson's Paradox via A Kernel-Based Partition Algorithm](https://mlsanity.com/paper/2603.02174/) (2026)
- [The Pragmatic Frames of Spurious Correlations in Machine Learning: Interpreting How and Why They Matter](https://mlsanity.com/paper/2411.04696/) (2024)
- [CausalSent: Interpretable Sentiment Classification with RieszNet](https://mlsanity.com/paper/2508.17576/) (2025)
- [Whitening Improves Robustness to Spurious Correlations in Linear Probes](https://mlsanity.com/paper/2609.39177/) (2026)

## Abstract (arXiv)

The term `spurious correlations' has been used in NLP to informally denote any undesirable feature-label correlations. However, a correlation can be undesirable because (i) the feature is irrelevant to the label (e.g. punctuation in a review), or (ii) the feature's effect on the label depends on the context (e.g. negation words in a review), which is ubiquitous in language tasks. In case (i), we want the model to be invariant to the feature, which is neither necessary nor sufficient for prediction. But in case (ii), even an ideal model (e.g. humans) must rely on the feature, since it is necessary (but not sufficient) for prediction. Therefore, a more fine-grained treatment of spurious features is needed to specify the desired model behavior. We formalize this distinction using a causal model and probabilities of necessity and sufficiency, which delineates the causal relations between a feature and a label. We then show that this distinction helps explain results of existing debiasing methods on different spurious features, and demystifies surprising results such as the encoding of spurious features in model representations after debiasing.