# Can Rationalization Improve Robustness?

arXiv: https://arxiv.org/abs/2204.11790 · HTML: https://mlsanity.com/paper/2204.11790/

## Key facts

- Cited by 36 papers in the map (data through 14 Oct 2026).
- Builds on 34 of its 47 references that are in the map.
- First posted on arXiv 25 Apr 2022.

## Builds on (most cited first, 30 of 34)

- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- ["Why Should I Trust You?": Explaining the Predictions of Any Classifier](https://mlsanity.com/paper/1602.04938/) (2016)
- [SQuAD: 100,000+ Questions for Machine Comprehension of Text](https://mlsanity.com/paper/1606.05250/) (2016)
- [Pointer Sentinel Mixture Models](https://mlsanity.com/paper/1609.07843/) (2016)
- [Categorical Reparameterization with Gumbel-Softmax](https://mlsanity.com/paper/1611.01144/) (2016)
- [Deep Variational Information Bottleneck](https://mlsanity.com/paper/1612.00410/) (2016)
- [Attention is not Explanation](https://mlsanity.com/paper/1902.10186/) (2019)
- [Universal Adversarial Triggers for Attacking and Analyzing NLP](https://mlsanity.com/paper/1908.07125/) (2019)
- [Towards Faithfully Interpretable NLP Systems: How should we define and evaluate faithfulness?](https://mlsanity.com/paper/2004.03685/) (2020)
- [Adversarial Examples for Evaluating Reading Comprehension Systems](https://mlsanity.com/paper/1707.07328/) (2017)
- [ERASER: A Benchmark to Evaluate Rationalized NLP Models](https://mlsanity.com/paper/1911.03429/) (2019)
- [Fooling LIME and SHAP: Adversarial Attacks on Post hoc Explanation Methods](https://mlsanity.com/paper/1911.02508/) (2019)
- [Learning the Difference that Makes a Difference with Counterfactually-Augmented Data](https://mlsanity.com/paper/1909.12434/) (2019)
- [Is Attention Interpretable?](https://mlsanity.com/paper/1906.03731/) (2019)
- [On the Robustness of Interpretability Methods](https://mlsanity.com/paper/1806.08049/) (2018)
- [Synthetic and Natural Noise Both Break Neural Machine Translation](https://mlsanity.com/paper/1711.02173/) (2017)
- [Adversarial Example Generation with Syntactically Controlled Paraphrase Networks](https://mlsanity.com/paper/1804.06059/) (2018)
- [Rationalizing Neural Predictions](https://mlsanity.com/paper/1606.04155/) (2016)
- [Measuring Association Between Labels and Free-Text Rationales](https://mlsanity.com/paper/2010.12762/) (2020)
- [WT5?! Training Text-to-Text Models to Explain their Predictions](https://mlsanity.com/paper/2004.14546/) (2020)
- [Invariant Rationalization](https://mlsanity.com/paper/2003.09772/) (2020)
- [Interpretable Neural Predictions with Differentiable Binary Variables](https://mlsanity.com/paper/1905.08160/) (2019)
- [Learning to Faithfully Rationalize by Construction](https://mlsanity.com/paper/2005.00115/) (2020)
- [Do Feature Attribution Methods Correctly Attribute Features?](https://mlsanity.com/paper/2104.14403/) (2021)
- [Rethinking Cooperative Rationalization: Introspective Extraction and Complement Control](https://mlsanity.com/paper/1910.13294/) (2019)
- [Look at the First Sentence: Position Bias in Question Answering](https://mlsanity.com/paper/2004.14602/) (2020)
- [Make Up Your Mind! Adversarial Generation of Inconsistent Natural Language Explanations](https://mlsanity.com/paper/1910.03065/) (2019)
- [An Information Bottleneck Approach for Controlling Conciseness in Rationale Extraction](https://mlsanity.com/paper/2005.00652/) (2020)
- [A causal framework for explaining the predictions of black-box sequence-to-sequence models](https://mlsanity.com/paper/1707.01943/) (2017)
- [Rationales for Sequential Predictions](https://mlsanity.com/paper/2109.06387/) (2021)

## Built on by (most cited first, 30 of 36)

- [Language Models Don't Always Say What They Think: Unfaithful Explanations in Chain-of-Thought Prompting](https://mlsanity.com/paper/2305.04388/) (2023)
- [The Unreliability of Explanations in Few-shot Prompting for Textual Reasoning](https://mlsanity.com/paper/2205.03401/) (2022)
- [Towards Faithful Model Explanation in NLP: A Survey](https://mlsanity.com/paper/2209.11326/) (2022)
- [InstructRAG: Instructing Retrieval-Augmented Generation via Self-Synthesized Rationales](https://mlsanity.com/paper/2406.13629/) (2024)
- [Towards Trustworthy and Aligned Machine Learning: A Data-centric Survey with Causality Perspectives](https://mlsanity.com/paper/2307.16851/) (2023)
- [How Interpretable are Reasoning Explanations from Prompting Large Language Models?](https://mlsanity.com/paper/2402.11863/) (2024)
- [Explanation-based Finetuning Makes Models More Robust to Spurious Cues](https://mlsanity.com/paper/2305.04990/) (2023)
- [CREST: A Joint Framework for Rationalization and Counterfactual Text Generation](https://mlsanity.com/paper/2305.17075/) (2023)
- [Discovering Spatio-Temporal Rationales for Video Question Answering](https://mlsanity.com/paper/2307.12058/) (2023)
- [D-Separation for Causal Self-Explanation](https://mlsanity.com/paper/2309.13391/) (2023)
- [Towards Trustworthy Explanation: On Causal Rationalization](https://mlsanity.com/paper/2306.14115/) (2023)
- [Enhancing the Rationale-Input Alignment for Self-explaining Rationalization](https://mlsanity.com/paper/2312.04103/) (2023)
- [Decoupled Rationalization with Asymmetric Learning Rates: A Flexible Lipschitz Restraint](https://mlsanity.com/paper/2305.13599/) (2023)
- [Rationale-Enhanced Language Models are Better Continual Relation Learners](https://mlsanity.com/paper/2310.06547/) (2023)
- [MGR: Multi-generator Based Rationalization](https://mlsanity.com/paper/2305.04492/) (2023)
- [Meta-Thinking in LLMs via Multi-Agent Reinforcement Learning: A Survey](https://mlsanity.com/paper/2504.14520/) (2025)
- [A.I. Robustness: a Human-Centered Perspective on Technological Challenges and Opportunities](https://mlsanity.com/paper/2210.08906/) (2022)
- [Does Self-Rationalization Improve Robustness to Spurious Correlations?](https://mlsanity.com/paper/2210.13575/) (2022)
- [Machine Learning Robustness: A Primer](https://mlsanity.com/paper/2404.00897/) (2024)
- [Leveraging Large Language Models for Bengali Math Word Problem Solving with Chain of Thought Reasoning](https://mlsanity.com/paper/2505.21354/) (2025)
- [A Survey of Computer Vision Technologies In Urban and Controlled-environment Agriculture](https://mlsanity.com/paper/2210.11318/) (2022)
- [Breaking Free from MMI: A New Frontier in Rationalization by Probing Input Utilization](https://mlsanity.com/paper/2503.06202/) (2025)
- [Adversarial Cooperative Rationalization: The Risk of Spurious Correlations in Even Clean Datasets](https://mlsanity.com/paper/2505.02118/) (2025)
- [A Survey on Measuring and Mitigating Reasoning Shortcuts in Machine Reading Comprehension](https://mlsanity.com/paper/2209.01824/) (2022)
- [The Impacts of Unanswerable Questions on the Robustness of Machine Reading Comprehension Models](https://mlsanity.com/paper/2302.00094/) (2023)
- [Rethinking Document-Level Relation Extraction: A Reality Check](https://mlsanity.com/paper/2306.08953/) (2023)
- [Learning to Ignore Adversarial Attacks](https://mlsanity.com/paper/2205.11551/) (2022)
- [Towards Robustness of Text-to-Visualization Translation against Lexical and Phrasal Variability](https://mlsanity.com/paper/2404.07135/) (2024)
- [Boosting Explainability through Selective Rationalization in Pre-trained Language Models](https://mlsanity.com/paper/2501.03182/) (2025)
- [Single-Sentence Reader: A Novel Approach for Addressing Answer Position Bias](https://mlsanity.com/paper/2308.04566/) (2023)

## Abstract (arXiv)

A growing line of work has investigated the development of neural NLP models that can produce rationales--subsets of input that can explain their model predictions. In this paper, we ask whether such rationale models can also provide robustness to adversarial attacks in addition to their interpretable nature. Since these models need to first generate rationales ("rationalizer") before making predictions ("predictor"), they have the potential to ignore noise or adversarially added text by simply masking it out of the generated rationale. To this end, we systematically generate various types of 'AddText' attacks for both token and sentence-level rationalization tasks, and perform an extensive empirical evaluation of state-of-the-art rationale models across five different tasks. Our experiments reveal that the rationale models show the promise to improve robustness, while they struggle in certain scenarios--when the rationalizer is sensitive to positional bias or lexical choices of attack text. Further, leveraging human rationale as supervision does not always translate to better performance. Our study is a first step towards exploring the interplay between interpretability and robustness in the rationalize-then-predict framework.