# Robustness to Spurious Correlations via Human Annotations

arXiv: https://arxiv.org/abs/2007.06661 · HTML: https://mlsanity.com/paper/2007.06661/

## Key facts

- Cited by 62 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 13 Jul 2020.

## Built on by (most cited first, 30 of 62)

- [Measure and Improve Robustness in NLP Models: A Survey](https://mlsanity.com/paper/2112.08313/) (2021)
- [Unveiling Transformers with LEGO: a synthetic reasoning task](https://mlsanity.com/paper/2206.04301/) (2022)
- [Eliciting Human Preferences with Language Models](https://mlsanity.com/paper/2310.11589/) (2023)
- [Distilling Model Failures as Directions in Latent Space](https://mlsanity.com/paper/2206.14754/) (2022)
- [Identifying and Mitigating Spurious Correlations for Improving Robustness in NLP Models](https://mlsanity.com/paper/2110.07736/) (2021)
- [GLUE-X: Evaluating Natural Language Understanding Models from an Out-of-distribution Generalization Perspective](https://mlsanity.com/paper/2211.08073/) (2022)
- [ZIN: When and How to Learn Invariance Without Environment Partition?](https://mlsanity.com/paper/2203.05818/) (2022)
- [Near-Optimal Linear Regression under Distribution Shift](https://mlsanity.com/paper/2106.12108/) (2021)
- [The Clever Hans Mirage: A Comprehensive Survey on Spurious Correlations in Machine Learning](https://mlsanity.com/paper/2402.12715/) (2024)
- [Invariant Feature Learning for Generalized Long-Tailed Classification](https://mlsanity.com/paper/2207.09504/) (2022)
- [Active Learning Helps Pretrained Models Learn the Intended Task](https://mlsanity.com/paper/2204.08491/) (2022)
- [Exploring the Efficacy of Automatically Generated Counterfactuals for Sentiment Analysis](https://mlsanity.com/paper/2106.15231/) (2021)
- [Understanding Rare Spurious Correlations in Neural Networks](https://mlsanity.com/paper/2202.05189/) (2022)
- [A Rationale-Centric Framework for Human-in-the-loop Machine Learning](https://mlsanity.com/paper/2203.12918/) (2022)
- [What to Learn, and How: Toward Effective Learning from Rationales](https://mlsanity.com/paper/2112.00071/) (2021)
- [NeuroCounterfactuals: Beyond Minimal-Edit Counterfactuals for Richer Data Augmentation](https://mlsanity.com/paper/2210.12365/) (2022)
- [Uncovering Main Causalities for Long-tailed Information Extraction](https://mlsanity.com/paper/2109.05213/) (2021)
- [Regularizing towards Causal Invariance: Linear Models with Proxies](https://mlsanity.com/paper/2103.02477/) (2021)
- [Robust Semantic Segmentation with Superpixel-Mix](https://mlsanity.com/paper/2108.00968/) (2021)
- [Synthesizing Adversarial Negative Responses for Robust Response Ranking and Evaluation](https://mlsanity.com/paper/2106.05894/) (2021)
- [Targeted Data Generation: Finding and Fixing Model Weaknesses](https://mlsanity.com/paper/2305.17804/) (2023)
- [Domain Generalization -- A Causal Perspective](https://mlsanity.com/paper/2209.15177/) (2022)
- [Towards Fine-grained Causal Reasoning and QA](https://mlsanity.com/paper/2204.07408/) (2022)
- [Enhancing Model Robustness and Fairness with Causality: A Regularization Approach](https://mlsanity.com/paper/2110.00911/) (2021)
- [Do Concept Bottleneck Models Respect Localities?](https://mlsanity.com/paper/2401.01259/) (2024)
- [FedGen: Generalizable Federated Learning for Sequential Data](https://mlsanity.com/paper/2211.01914/) (2022)
- [Feature Dropout: Revisiting the Role of Augmentations in Contrastive Learning](https://mlsanity.com/paper/2212.08378/) (2022)
- [Out of Spuriousity: Improving Robustness to Spurious Correlations without Group Annotations](https://mlsanity.com/paper/2407.14974/) (2024)
- [Adversarial Self-Attention for Language Understanding](https://mlsanity.com/paper/2206.12608/) (2022)
- [Probing Spurious Correlations in Popular Event-Based Rumor Detection Benchmarks](https://mlsanity.com/paper/2209.08799/) (2022)

## Abstract (arXiv)

The reliability of machine learning systems critically assumes that the associations between features and labels remain similar between training and test distributions. However, unmeasured variables, such as confounders, break this assumption---useful correlations between features and labels at training time can become useless or even harmful at test time. For example, high obesity is generally predictive for heart disease, but this relation may not hold for smokers who generally have lower rates of obesity and higher rates of heart disease. We present a framework for making models robust to spurious correlations by leveraging humans' common sense knowledge of causality. Specifically, we use human annotation to augment each training example with a potential unmeasured variable (i.e. an underweight patient with heart disease may be a smoker), reducing the problem to a covariate shift problem. We then introduce a new distributionally robust optimization objective over unmeasured variables (UV-DRO) to control the worst-case loss over possible test-time shifts. Empirically, we show improvements of 5-10% on a digit recognition task confounded by rotation, and 1.5-5% on the task of analyzing NYPD Police Stops confounded by location.