# Exposing Shallow Heuristics of Relation Extraction Models with Challenge Data

arXiv: https://arxiv.org/abs/2010.03656 · HTML: https://mlsanity.com/paper/2010.03656/

## Key facts

- Cited by 20 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 7 Oct 2020.

## Built on by (most cited first, 20 of 20)

- [Are NLP Models really able to Solve Simple Math Word Problems?](https://mlsanity.com/paper/2103.07191/) (2021)
- [What You See is What You Read? Improving Text-Image Alignment Evaluation](https://mlsanity.com/paper/2305.10400/) (2023)
- [State-of-the-art generalisation research in NLP: A taxonomy and review](https://mlsanity.com/paper/2210.03050/) (2022)
- ["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)
- [Automatic Generation of Contrast Sets from Scene Graphs: Probing the Compositional Consistency of GQA](https://mlsanity.com/paper/2103.09591/) (2021)
- [Diagnosing AI Explanation Methods with Folk Concepts of Behavior](https://mlsanity.com/paper/2201.11239/) (2022)
- [Finding Dataset Shortcuts with Grammar Induction](https://mlsanity.com/paper/2210.11560/) (2022)
- [A Dataset for N-ary Relation Extraction of Drug Combinations](https://mlsanity.com/paper/2205.02289/) (2022)
- [Fighting Bias with Bias: Promoting Model Robustness by Amplifying Dataset Biases](https://mlsanity.com/paper/2305.18917/) (2023)
- [Linguistic representations for fewer-shot relation extraction across domains](https://mlsanity.com/paper/2307.03823/) (2023)
- [Quantifying the Task-Specific Information in Text-Based Classifications](https://mlsanity.com/paper/2110.08931/) (2021)
- [Separating Retention from Extraction in the Evaluation of End-to-end Relation Extraction](https://mlsanity.com/paper/2109.12008/) (2021)
- [MedDistant19: Towards an Accurate Benchmark for Broad-Coverage Biomedical Relation Extraction](https://mlsanity.com/paper/2204.04779/) (2022)
- [Understanding Text Classification Data and Models Using Aggregated Input Salience](https://mlsanity.com/paper/2211.05485/) (2022)
- [Pointing out the Shortcomings of Relation Extraction Models with Semantically Motivated Adversarials](https://mlsanity.com/paper/2402.19076/) (2024)
- [Improving Span Representation for Domain-adapted Coreference Resolution](https://mlsanity.com/paper/2109.09811/) (2021)
- [FREDA: Flexible Relation Extraction Data Annotation](https://mlsanity.com/paper/2204.07150/) (2022)
- [CHARP: Conversation History AwaReness Probing for Knowledge-grounded Dialogue Systems](https://mlsanity.com/paper/2405.15110/) (2024)
- [Neuro-Symbolic Contrastive Learning for Cross-domain Inference](https://mlsanity.com/paper/2502.09213/) (2025)
- [Do Large Language Models know who did what to whom?](https://mlsanity.com/paper/2504.16884/) (2025)

## Abstract (arXiv)

The process of collecting and annotating training data may introduce distribution artifacts which may limit the ability of models to learn correct generalization behavior. We identify failure modes of SOTA relation extraction (RE) models trained on TACRED, which we attribute to limitations in the data annotation process. We collect and annotate a challenge-set we call Challenging RE (CRE), based on naturally occurring corpus examples, to benchmark this behavior. Our experiments with four state-of-the-art RE models show that they have indeed adopted shallow heuristics that do not generalize to the challenge-set data. Further, we find that alternative question answering modeling performs significantly better than the SOTA models on the challenge-set, despite worse overall TACRED performance. By adding some of the challenge data as training examples, the performance of the model improves. Finally, we provide concrete suggestion on how to improve RE data collection to alleviate this behavior.