# Probing Neural Network Comprehension of Natural Language Arguments

arXiv: https://arxiv.org/abs/1907.07355 · HTML: https://mlsanity.com/paper/1907.07355/

## Key facts

- Cited by 214 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 17 Jul 2019.

## Built on by (most cited first, 30 of 214)

- [Pre-train, Prompt, and Predict: A Systematic Survey of Prompting Methods in Natural Language Processing](https://mlsanity.com/paper/2107.13586/) (2021)
- [AI and the Everything in the Whole Wide World Benchmark](https://mlsanity.com/paper/2111.15366/) (2021)
- [Pre-Trained Models: Past, Present and Future](https://mlsanity.com/paper/2106.07139/) (2021)
- [What Will it Take to Fix Benchmarking in Natural Language Understanding?](https://mlsanity.com/paper/2104.02145/) (2021)
- [Partial success in closing the gap between human and machine vision](https://mlsanity.com/paper/2106.07411/) (2021)
- [On Feature Learning in the Presence of Spurious Correlations](https://mlsanity.com/paper/2210.11369/) (2022)
- [mGPT: Few-Shot Learners Go Multilingual](https://mlsanity.com/paper/2204.07580/) (2022)
- [The Debate Over Understanding in AI's Large Language Models](https://mlsanity.com/paper/2210.13966/) (2022)
- [Toward Transparent AI: A Survey on Interpreting the Inner Structures of Deep Neural Networks](https://mlsanity.com/paper/2207.13243/) (2022)
- [BERT: A Review of Applications in Natural Language Processing and Understanding](https://mlsanity.com/paper/2103.11943/) (2021)
- [An Empirical Study of LLM-as-a-Judge for LLM Evaluation: Fine-tuned Judge Model is not a General Substitute for GPT-4](https://mlsanity.com/paper/2403.02839/) (2024)
- [Shortcut Learning of Large Language Models in Natural Language Understanding](https://mlsanity.com/paper/2208.11857/) (2022)
- [Towards Interpreting and Mitigating Shortcut Learning Behavior of NLU Models](https://mlsanity.com/paper/2103.06922/) (2021)
- [GLUE-X: Evaluating Natural Language Understanding Models from an Out-of-distribution Generalization Perspective](https://mlsanity.com/paper/2211.08073/) (2022)
- [Probable Domain Generalization via Quantile Risk Minimization](https://mlsanity.com/paper/2207.09944/) (2022)
- [Fairness via Representation Neutralization](https://mlsanity.com/paper/2106.12674/) (2021)
- [Availability Attacks Create Shortcuts](https://mlsanity.com/paper/2111.00898/) (2021)
- [Do Large Language Models know what humans know?](https://mlsanity.com/paper/2209.01515/) (2022)
- [Comparing Test Sets with Item Response Theory](https://mlsanity.com/paper/2106.00840/) (2021)
- [Towards Last-layer Retraining for Group Robustness with Fewer Annotations](https://mlsanity.com/paper/2309.08534/) (2023)
- [Multilingual HateCheck: Functional Tests for Multilingual Hate Speech Detection Models](https://mlsanity.com/paper/2206.09917/) (2022)
- [Large Language Models Hallucination: A Comprehensive Survey](https://mlsanity.com/paper/2510.06265/) (2025)
- [Event knowledge in large language models: the gap between the impossible and the unlikely](https://mlsanity.com/paper/2212.01488/) (2022)
- [InjecGuard: Benchmarking and Mitigating Over-defense in Prompt Injection Guardrail Models](https://mlsanity.com/paper/2410.22770/) (2024)
- [Attacking Open-domain Question Answering by Injecting Misinformation](https://mlsanity.com/paper/2110.07803/) (2021)
- [GPT-4 Can't Reason](https://mlsanity.com/paper/2308.03762/) (2023)
- [The Dangers of Underclaiming: Reasons for Caution When Reporting How NLP Systems Fail](https://mlsanity.com/paper/2110.08300/) (2021)
- [KOBEST: Korean Balanced Evaluation of Significant Tasks](https://mlsanity.com/paper/2204.04541/) (2022)
- [Finding and Editing Multi-Modal Neurons in Pre-Trained Transformers](https://mlsanity.com/paper/2311.07470/) (2023)
- [Modular and Parameter-Efficient Multimodal Fusion with Prompting](https://mlsanity.com/paper/2203.08055/) (2022)

## Abstract (arXiv)

We are surprised to find that BERT's peak performance of 77% on the Argument Reasoning Comprehension Task reaches just three points below the average untrained human baseline. However, we show that this result is entirely accounted for by exploitation of spurious statistical cues in the dataset. We analyze the nature of these cues and demonstrate that a range of models all exploit them. This analysis informs the construction of an adversarial dataset on which all models achieve random accuracy. Our adversarial dataset provides a more robust assessment of argument comprehension and should be adopted as the standard in future work.