# Explanation-based Finetuning Makes Models More Robust to Spurious Cues

arXiv: https://arxiv.org/abs/2305.04990 · HTML: https://mlsanity.com/paper/2305.04990/

## Key facts

- Cited by 25 papers in the map (data through 14 Oct 2026).
- Builds on 33 of its 41 references that are in the map.
- First posted on arXiv 8 May 2023.

## Builds on (most cited first, 30 of 33)

- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer](https://mlsanity.com/paper/1910.10683/) (2019)
- [Chain-of-Thought Prompting Elicits Reasoning in Large Language Models](https://mlsanity.com/paper/2201.11903/) (2022)
- [Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks](https://mlsanity.com/paper/1908.10084/) (2019)
- [BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehension](https://mlsanity.com/paper/1910.13461/) (2019)
- [OPT: Open Pre-trained Transformer Language Models](https://mlsanity.com/paper/2205.01068/) (2022)
- [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)
- [Annotation Artifacts in Natural Language Inference Data](https://mlsanity.com/paper/1803.02324/) (2018)
- [e-SNLI: Natural Language Inference with Natural Language Explanations](https://mlsanity.com/paper/1812.01193/) (2018)
- [Learning the Difference that Makes a Difference with Counterfactually-Augmented Data](https://mlsanity.com/paper/1909.12434/) (2019)
- [Social Bias Frames: Reasoning about Social and Power Implications of Language](https://mlsanity.com/paper/1911.03891/) (2019)
- [Hypothesis Only Baselines in Natural Language Inference](https://mlsanity.com/paper/1805.01042/) (2018)
- [Can language models learn from explanations in context?](https://mlsanity.com/paper/2204.02329/) (2022)
- [Examining Gender and Race Bias in Two Hundred Sentiment Analysis Systems](https://mlsanity.com/paper/1805.04508/) (2018)
- [Towards Understanding Chain-of-Thought Prompting: An Empirical Study of What Matters](https://mlsanity.com/paper/2212.10001/) (2022)
- [Gender Bias in Neural Natural Language Processing](https://mlsanity.com/paper/1807.11714/) (2018)
- [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)
- [The Unreliability of Explanations in Few-shot Prompting for Textual Reasoning](https://mlsanity.com/paper/2205.03401/) (2022)
- [Measuring Association Between Labels and Free-Text Rationales](https://mlsanity.com/paper/2010.12762/) (2020)
- [End-to-End Bias Mitigation by Modelling Biases in Corpora](https://mlsanity.com/paper/1909.06321/) (2019)
- [How Much Reading Does Reading Comprehension Require? A Critical Investigation of Popular Benchmarks](https://mlsanity.com/paper/1808.04926/) (2018)
- [CREAK: A Dataset for Commonsense Reasoning over Entity Knowledge](https://mlsanity.com/paper/2109.01653/) (2021)
- [Learning from others' mistakes: Avoiding dataset biases without modeling them](https://mlsanity.com/paper/2012.01300/) (2020)
- [Few-Shot Self-Rationalization with Natural Language Prompts](https://mlsanity.com/paper/2111.08284/) (2021)
- [A Thorough Examination of the CNN/Daily Mail Reading Comprehension Task](https://mlsanity.com/paper/1606.02858/) (2016)
- [Generating Data to Mitigate Spurious Correlations in Natural Language Inference Datasets](https://mlsanity.com/paper/2203.12942/) (2022)
- [Does It Make Sense? And Why? A Pilot Study for Sense Making and Explanation](https://mlsanity.com/paper/1906.00363/) (2019)
- [Can Rationalization Improve Robustness?](https://mlsanity.com/paper/2204.11790/) (2022)
- [Supervising Model Attention with Human Explanations for Robust Natural Language Inference](https://mlsanity.com/paper/2104.08142/) (2021)

## Built on by (most cited first, 25 of 25)

- [The Clever Hans Mirage: A Comprehensive Survey on Spurious Correlations in Machine Learning](https://mlsanity.com/paper/2402.12715/) (2024)
- [HARE: Explainable Hate Speech Detection with Step-by-Step Reasoning](https://mlsanity.com/paper/2311.00321/) (2023)
- [AI-Slop to AI-Polish? Aligning Language Models through Edit-Based Writing Rewards and Test-time Computation](https://mlsanity.com/paper/2504.07532/) (2025)
- [Small Language Models are Good Too: An Empirical Study of Zero-Shot Classification](https://mlsanity.com/paper/2404.11122/) (2024)
- [Understanding and Mitigating Spurious Correlations in Text Classification with Neighborhood Analysis](https://mlsanity.com/paper/2305.13654/) (2023)
- [ALMANACS: A Simulatability Benchmark for Language Model Explainability](https://mlsanity.com/paper/2312.12747/) (2023)
- [EmoDynamiX: Emotional Support Dialogue Strategy Prediction by Modelling MiXed Emotions and Discourse Dynamics](https://mlsanity.com/paper/2408.08782/) (2024)
- [STRUX: An LLM for Decision-Making with Structured Explanations](https://mlsanity.com/paper/2410.12583/) (2024)
- [Fast-Slow-Thinking: Complex Task Solving with Large Language Models](https://mlsanity.com/paper/2504.08690/) (2025)
- [ICLEF: In-Context Learning with Expert Feedback for Explainable Style Transfer](https://mlsanity.com/paper/2309.08583/) (2023)
- [EAVIT: Efficient and Accurate Human Value Identification from Text data via LLMs](https://mlsanity.com/paper/2505.12792/) (2025)
- [Explanation-aware Soft Ensemble Empowers Large Language Model In-context Learning](https://mlsanity.com/paper/2311.07099/) (2023)
- [Rescue: Ranking LLM Responses with Partial Ordering to Improve Response Generation](https://mlsanity.com/paper/2311.09136/) (2023)
- [RORA: Robust Free-Text Rationale Evaluation](https://mlsanity.com/paper/2402.18678/) (2024)
- [Improving the OOD Performance of Closed-Source LLMs on NLI Through Strategic Data Selection](https://mlsanity.com/paper/2505.20209/) (2025)
- [Explaining Relationships Among Research Papers](https://mlsanity.com/paper/2402.13426/) (2024)
- [Using Natural Language Explanations to Improve Robustness of In-context Learning](https://mlsanity.com/paper/2311.07556/) (2023)
- [Social Human Robot Embodied Conversation (SHREC) Dataset: Benchmarking Foundational Models' Social Reasoning](https://mlsanity.com/paper/2504.13898/) (2025)
- [Exploring Explanations Improves the Robustness of In-Context Learning](https://mlsanity.com/paper/2506.02378/) (2025)
- [Single-weight Model Editing for Post-hoc Spurious Correlation Neutralization](https://mlsanity.com/paper/2501.14182/) (2025)
- [Corrections Meet Explanations: A Unified Framework for Explainable Grammatical Error Correction](https://mlsanity.com/paper/2502.15261/) (2025)
- [Robustness of LLM-enabled vehicle trajectory prediction under data security threats](https://mlsanity.com/paper/2511.13753/) (2025)
- [Supervised Fine-Tuning Needs to Unlock the Potential of Token Priority](https://mlsanity.com/paper/2602.01227/) (2026)
- [Improving Implicit Discourse Relation Recognition with Natural Language Explanations from LLMs](https://mlsanity.com/paper/2602.21763/) (2026)
- [From Expression to Reaction: Role-aware Visual Transfer and Stimulus-guided Reasoning for Interlocutor Emotion Recognition](https://mlsanity.com/paper/2610.03016/) (2026)

## Abstract (arXiv)

Large Language Models (LLMs) are so powerful that they sometimes learn correlations between labels and features that are irrelevant to the task, leading to poor generalization on out-of-distribution data. We propose explanation-based finetuning as a general approach to mitigate LLMs' reliance on spurious correlations. Unlike standard finetuning where the model only predicts the answer given the input, we finetune the model to additionally generate a free-text explanation supporting its answer. To evaluate our method, we finetune the model on artificially constructed training sets containing different types of spurious cues, and test it on a test set without these cues. Compared to standard finetuning, our method makes GPT-3 (davinci) remarkably more robust against spurious cues in terms of accuracy drop across four classification tasks: ComVE (+1.2), CREAK (+9.1), e-SNLI (+15.4), and SBIC (+6.5). The efficacy generalizes across multiple model families and scales, with greater gains for larger models. Finally, our method also works well with explanations generated by the model, implying its applicability to more datasets without human-written explanations.