mlsanityData to 7 Oct 2026

Probing Neural Network Comprehension of Natural Language Arguments

arXiv 1907.07355 ↗First posted 17 Jul 2019cs.CLcited by 212

Not on the map. Why?

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.

Key facts. Cited by 212 papers in the map (data through 7 Oct 2026). First posted on arXiv 17 Jul 2019.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 53; 2022: 65; 2023: 35; 2024: 29; 2025: 12; 2026: 18.

212 papers.

Sort:

By area:

  1. 2021Pre-train, Prompt, and Predict: A Systematic Survey of Prompting Methods in Natural Language Processingcited by 2,222
  2. 2021AI and the Everything in the Whole Wide World Benchmarkcited by 340
  3. 2021Pre-Trained Models: Past, Present and Futurecited by 319
  4. 2021What Will it Take to Fix Benchmarking in Natural Language Understanding?cited by 200
  5. 2021Partial success in closing the gap between human and machine visioncited by 185
  6. 2022On Feature Learning in the Presence of Spurious Correlationscited by 145
  7. 2022The Debate Over Understanding in AI's Large Language Modelscited by 142
  8. 2022mGPT: Few-Shot Learners Go Multilingualcited by 141
  9. 2022Toward Transparent AI: A Survey on Interpreting the Inner Structures of Deep Neural Networkscited by 126
  10. 2021BERT: A Review of Applications in Natural Language Processing and Understandingcited by 109
  11. 2024An Empirical Study of LLM-as-a-Judge for LLM Evaluation: Fine-tuned Judge Model is not a General Substitute for GPT-4cited by 90
  12. 2022Shortcut Learning of Large Language Models in Natural Language Understandingcited by 83
38 more
  1. 2021Towards Interpreting and Mitigating Shortcut Learning Behavior of NLU Modelscited by 78
  2. 2022GLUE-X: Evaluating Natural Language Understanding Models from an Out-of-distribution Generalization Perspectivecited by 72
  3. 2022Probable Domain Generalization via Quantile Risk Minimizationcited by 66
  4. 2021Fairness via Representation Neutralizationcited by 63
  5. 2021Availability Attacks Create Shortcutscited by 60
  6. 2022Do Large Language Models know what humans know?cited by 60
  7. 2021Comparing Test Sets with Item Response Theorycited by 55
  8. 2023Towards Last-layer Retraining for Group Robustness with Fewer Annotationscited by 54
  9. 2022Multilingual HateCheck: Functional Tests for Multilingual Hate Speech Detection Modelscited by 51
  10. 2025Large Language Models Hallucination: A Comprehensive Surveycited by 45
  11. 2022Event knowledge in large language models: the gap between the impossible and the unlikelycited by 44
  12. 2021Attacking Open-domain Question Answering by Injecting Misinformationcited by 41
  13. 2023GPT-4 Can't Reasoncited by 39
  14. 2021The Dangers of Underclaiming: Reasons for Caution When Reporting How NLP Systems Failcited by 38
  15. 2022KOBEST: Korean Balanced Evaluation of Significant Taskscited by 37
  16. 2023Finding and Editing Multi-Modal Neurons in Pre-Trained Transformerscited by 36
  17. 2024InjecGuard: Benchmarking and Mitigating Over-defense in Prompt Injection Guardrail Modelscited by 36
  18. 2022Modular and Parameter-Efficient Multimodal Fusion with Promptingcited by 35
  19. 2023Towards LogiGLUE: A Brief Survey and A Benchmark for Analyzing Logical Reasoning Capabilities of Language Modelscited by 35
  20. 2021Knowledge Enhanced Pretrained Language Models: A Compreshensive Surveycited by 33
  21. 2022CoCo: Coherence-Enhanced Machine-Generated Text Detection Under Data Limitation With Contrastive Learningcited by 32
  22. 2023Mind the instructions: a holistic evaluation of consistency and interactions in prompt-based learningcited by 31
  23. 2024On the Challenges and Opportunities in Generative AIcited by 31
  24. 2024Examining the robustness of LLM evaluation to the distributional assumptions of benchmarkscited by 31
  25. 2021Do Syntactic Probes Probe Syntax? Experiments with Jabberwocky Probingcited by 28
  26. 2023Counterfactual reasoning: Testing language models' understanding of hypothetical scenarioscited by 27
  27. 2022RuCoLA: Russian Corpus of Linguistic Acceptabilitycited by 26
  28. 2023Zero-shot Faithful Factual Error Correctioncited by 26
  29. 2023It Takes Two to Tango: Navigating Conceptualizations of NLP Tasks and Measurements of Performancecited by 25
  30. 2024The Clever Hans Effect in Unsupervised Learningcited by 24
  31. 2021Avoiding Inference Heuristics in Few-shot Prompt-based Finetuningcited by 23
  32. 2023GradientCoin: A Peer-to-Peer Decentralized Large Language Modelscited by 23
  33. 2023What's the Meaning of Superhuman Performance in Today's NLU?cited by 22
  34. 2022Eye-gaze-guided Vision Transformer for Rectifying Shortcut Learningcited by 20
  35. 2021Tiered Reasoning for Intuitive Physics: Toward Verifiable Commonsense Language Understandingcited by 19
  36. 2022Clinical Prompt Learning with Frozen Language Modelscited by 19
  37. 2023KILM: Knowledge Injection into Encoder-Decoder Language Modelscited by 19
  38. 2021ProtAugment: Unsupervised diverse short-texts paraphrasing for intent detection meta-learningcited by 18

50 most cited of 212 All 212 (JSON)