# DISCO: DISCovering Overfittings as Causal Rules for Text Classification Models

arXiv: https://arxiv.org/abs/2411.04649 · HTML: https://mlsanity.com/paper/2411.04649/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 23 of its 38 references that are in the map.
- First posted on arXiv 7 Nov 2024.

## Builds on (most cited first, 23 of 23)

- [Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks](https://mlsanity.com/paper/1908.10084/) (2019)
- [A Unified Approach to Interpreting Model Predictions](https://mlsanity.com/paper/1705.07874/) (2017)
- [Axiomatic Attribution for Deep Networks](https://mlsanity.com/paper/1703.01365/) (2017)
- [Shortcut Learning in Deep Neural Networks](https://mlsanity.com/paper/2004.07780/) (2020)
- [ERASER: A Benchmark to Evaluate Rationalized NLP Models](https://mlsanity.com/paper/1911.03429/) (2019)
- [Learning the Difference that Makes a Difference with Counterfactually-Augmented Data](https://mlsanity.com/paper/1909.12434/) (2019)
- [Rationalizing Neural Predictions](https://mlsanity.com/paper/1606.04155/) (2016)
- [Compositional Explanations of Neurons](https://mlsanity.com/paper/2006.14032/) (2020)
- [Post-hoc Interpretability for Neural NLP: A Survey](https://mlsanity.com/paper/2108.04840/) (2021)
- [Invariant Rationalization](https://mlsanity.com/paper/2003.09772/) (2020)
- [Interpretable Machine Learning -- A Brief History, State-of-the-Art and Challenges](https://mlsanity.com/paper/2010.09337/) (2020)
- [Identifying and Mitigating Spurious Correlations for Improving Robustness in NLP Models](https://mlsanity.com/paper/2110.07736/) (2021)
- ["Will You Find These Shortcuts?" A Protocol for Evaluating the Faithfulness of Input Salience Methods for Text Classification](https://mlsanity.com/paper/2111.07367/) (2021)
- [Inferring Which Medical Treatments Work from Reports of Clinical Trials](https://mlsanity.com/paper/1904.01606/) (2019)
- [Identifying Spurious Correlations for Robust Text Classification](https://mlsanity.com/paper/2010.02458/) (2020)
- [Explain and Predict, and then Predict Again](https://mlsanity.com/paper/2101.04109/) (2021)
- [Simplified Data Wrangling with ir_datasets](https://mlsanity.com/paper/2103.02280/) (2021)
- [A Rationale-Centric Framework for Human-in-the-loop Machine Learning](https://mlsanity.com/paper/2203.12918/) (2022)
- [Explainable Information Retrieval: A Survey](https://mlsanity.com/paper/2211.02405/) (2022)
- [The Change that Matters in Discourse Parsing: Estimating the Impact of Domain Shift on Parser Error](https://mlsanity.com/paper/2203.11317/) (2022)
- [Towards Benchmarking the Utility of Explanations for Model Debugging](https://mlsanity.com/paper/2105.04505/) (2021)
- [Interpreting Deep Learning Model Using Rule-based Method](https://mlsanity.com/paper/2010.07824/) (2020)
- [SparCAssist: A Model Risk Assessment Assistant Based on Sparse Generated Counterfactuals](https://mlsanity.com/paper/2205.01588/) (2022)

## Abstract (arXiv)

With the rapid advancement of neural language models, the deployment of over-parameterized models has surged, increasing the need for interpretable explanations comprehensible to human inspectors. Existing post-hoc interpretability methods, which often focus on unigram features of single input textual instances, fail to capture the models' decision-making process fully. Additionally, many methods do not differentiate between decisions based on spurious correlations and those based on a holistic understanding of the input. Our paper introduces DISCO, a novel method for discovering global, rule-based explanations by identifying causal n-gram associations with model predictions. This method employs a scalable sequence mining technique to extract relevant text spans from training data, associate them with model predictions, and conduct causality checks to distill robust rules that elucidate model behavior. These rules expose potential overfitting and provide insights into misleading feature combinations. We validate DISCO through extensive testing, demonstrating its superiority over existing methods in offering comprehensive insights into complex model behaviors. Our approach successfully identifies all shortcuts manually introduced into the training data (100% detection rate on the MultiRC dataset), resulting in an 18.8% regression in model performance -- a capability unmatched by any other method. Furthermore, DISCO supports interactive explanations, enabling human inspectors to distinguish spurious causes in the rule-based output. This alleviates the burden of abundant instance-wise explanations and helps assess the model's risk when encountering out-of-distribution (OOD) data.