# Label-Descriptive Patterns and Their Application to Characterizing Classification Errors

arXiv: https://arxiv.org/abs/2110.09599 · HTML: https://mlsanity.com/paper/2110.09599/

## Key facts

- Cited by 10 papers in the map (data through 14 Oct 2026).
- Builds on 15 of its 64 references that are in the map.
- First posted on arXiv 18 Oct 2021.

## Builds on (most cited first, 15 of 15)

- ["Why Should I Trust You?": Explaining the Predictions of Any Classifier](https://mlsanity.com/paper/1602.04938/) (2016)
- [LXMERT: Learning Cross-Modality Encoder Representations from Transformers](https://mlsanity.com/paper/1908.07490/) (2019)
- [Beyond Accuracy: Behavioral Testing of NLP models with CheckList](https://mlsanity.com/paper/2005.04118/) (2020)
- [End-to-end Sequence Labeling via Bi-directional LSTM-CNNs-CRF](https://mlsanity.com/paper/1603.01354/) (2016)
- [Visual7W: Grounded Question Answering in Images](https://mlsanity.com/paper/1511.03416/) (2015)
- [Learning Word Vectors for 157 Languages](https://mlsanity.com/paper/1802.06893/) (2018)
- [Distilling a Neural Network Into a Soft Decision Tree](https://mlsanity.com/paper/1711.09784/) (2017)
- [A Survey of the State of Explainable AI for Natural Language Processing](https://mlsanity.com/paper/2010.00711/) (2020)
- [Interpretable & Explorable Approximations of Black Box Models](https://mlsanity.com/paper/1707.01154/) (2017)
- [Transfer Learning and Distant Supervision for Multilingual Transformer Models: A Study on African Languages](https://mlsanity.com/paper/2010.03179/) (2020)
- [Interpretable multiclass classification by MDL-based rule lists](https://mlsanity.com/paper/1905.00328/) (2019)
- [Fast and Memory-Efficient Significant Pattern Mining via Permutation Testing](https://mlsanity.com/paper/1502.04315/) (2015)
- [C-SALT: Mining Class-Specific ALTerations in Boolean Matrix Factorization](https://mlsanity.com/paper/1906.09907/) (2019)
- [QADiver: Interactive Framework for Diagnosing QA Models](https://mlsanity.com/paper/1812.00161/) (2018)
- [Summarizing Data Succinctly with the Most Informative Itemsets](https://mlsanity.com/paper/1904.11134/) (2019)

## Built on by (most cited first, 10 of 10)

- [Persistent Personas? Role-Playing, Instruction Following, and Safety in Extended Interactions](https://mlsanity.com/paper/2512.12775/) (2025)
- [Intrinsic User-Centric Interpretability through Global Mixture of Experts](https://mlsanity.com/paper/2402.02933/) (2024)
- [Evaluating Robustness of Large Language Models Against Multilingual Typographical Errors](https://mlsanity.com/paper/2510.09536/) (2025)
- [Finding Interpretable Class-Specific Patterns through Efficient Neural Search](https://mlsanity.com/paper/2312.04311/) (2023)
- [Establishing Trustworthiness: Rethinking Tasks and Model Evaluation](https://mlsanity.com/paper/2310.05442/) (2023)
- [What's the Difference? Supporting Users in Identifying the Effects of Prompt and Model Changes Through Token Patterns](https://mlsanity.com/paper/2504.15815/) (2025)
- [CohEx: A Generalized Framework for Cohort Explanation](https://mlsanity.com/paper/2410.13190/) (2024)
- [Rule Mining for Correcting Classification Models](https://mlsanity.com/paper/2310.06446/) (2023)
- [Understanding and Mitigating Classification Errors Through Interpretable Token Patterns](https://mlsanity.com/paper/2311.10920/) (2023)
- [Interpreting Inflammation Prediction Model via Tag-based Cohort Explanation](https://mlsanity.com/paper/2410.14082/) (2024)

## Abstract (arXiv)

State-of-the-art deep learning methods achieve human-like performance on many tasks, but make errors nevertheless. Characterizing these errors in easily interpretable terms gives insight into whether a classifier is prone to making systematic errors, but also gives a way to act and improve the classifier. We propose to discover those feature-value combinations (i.e., patterns) that strongly correlate with correct resp. erroneous predictions to obtain a global and interpretable description for arbitrary classifiers. We show this is an instance of the more general label description problem, which we formulate in terms of the Minimum Description Length principle. To discover a good pattern set, we develop the efficient Premise algorithm. Through an extensive set of experiments we show it performs very well in practice on both synthetic and real-world data. Unlike existing solutions, it ably recovers ground truth patterns, even on highly imbalanced data over many features. Through two case studies on Visual Question Answering and Named Entity Recognition, we confirm that Premise gives clear and actionable insight into the systematic errors made by modern NLP classifiers.