mlsanityData to 7 Oct 2026

Adversarial Examples for Evaluating Reading Comprehension Systems

arXiv 1707.07328 ↗First posted 23 Jul 2017cs.CL cs.LGcited by 568

Not on the map. Why?

Cited by 28 in LLM safety and security, but placed by its own abstract, which didn’t fit.

Abstract (arXiv)

Standard accuracy metrics indicate that reading comprehension systems are making rapid progress, but the extent to which these systems truly understand language remains unclear. To reward systems with real language understanding abilities, we propose an adversarial evaluation scheme for the Stanford Question Answering Dataset (SQuAD). Our method tests whether systems can answer questions about paragraphs that contain adversarially inserted sentences, which are automatically generated to distract computer systems without changing the correct answer or misleading humans. In this adversarial setting, the accuracy of sixteen published models drops from an average of 75% F1 score to 36%; when the adversary is allowed to add ungrammatical sequences of words, average accuracy on four models decreases further to 7%. We hope our insights will motivate the development of new models that understand language more precisely.

Key facts. Cited by 568 papers in the map (data through 7 Oct 2026). First posted on arXiv 23 Jul 2017.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 156; 2022: 116; 2023: 109; 2024: 94; 2025: 57; 2026: 36.

568 papers.

Sort:

By area:

  1. 2022Robust Speech Recognition via Large-Scale Weak Supervisioncited by 3,972
  2. 2023Universal and Transferable Adversarial Attacks on Aligned Language Modelscited by 2,914
  3. 2022Red Teaming Language Models with Language Modelscited by 983
  4. 2021MuSiQue: Multihop Questions via Single-hop Question Compositioncited by 893
  5. 2023Large Language Models Can Be Easily Distracted by Irrelevant Contextcited by 820
  6. 2022Red Teaming Language Models to Reduce Harms: Methods, Scaling Behaviors, and Lessons Learnedcited by 702
  7. 2021Dynabench: Rethinking Benchmarking in NLPcited by 447
  8. 2023Making Retrieval-Augmented Language Models Robust to Irrelevant Contextcited by 364
  9. 2021Measuring and Improving Consistency in Pretrained Language Modelscited by 361
  10. 2023Are aligned neural networks adversarially aligned?cited by 305
  11. 2023Low-Resource Languages Jailbreak GPT-4cited by 290
  12. 2022Prompting GPT-3 To Be Reliablecited by 258
38 more
  1. 2022ASQA: Factoid Questions Meet Long-Form Answerscited by 231
  2. 2021Adversarial GLUE: A Multi-Task Benchmark for Robustness Evaluation of Language Modelscited by 227
  3. 2021What Will it Take to Fix Benchmarking in Natural Language Understanding?cited by 200
  4. 2022WANLI: Worker and AI Collaboration for Natural Language Inference Dataset Creationcited by 200
  5. 2023Survey of Vulnerabilities in Large Language Models Revealed by Adversarial Attackscited by 189
  6. 2023Automatically Auditing Large Language Models via Discrete Optimizationcited by 187
  7. 2023Adversarial Example Does Good: Preventing Painting Imitation from Diffusion Models via Adversarial Examplescited by 163
  8. 2023Poisoning Retrieval Corpora by Injecting Adversarial Passagescited by 153
  9. 2024Robust Prompt Optimization for Defending Language Models Against Jailbreaking Attackscited by 150
  10. 2021Mind the Style of Text! Adversarial and Backdoor Attacks Based on Text Style Transfercited by 147
  11. 2024LLM Defenses Are Not Robust to Multi-Turn Human Jailbreaks Yetcited by 145
  12. 2022Impact of Pretraining Term Frequencies on Few-Shot Reasoningcited by 142
  13. 2022Rich Knowledge Sources Bring Complex Knowledge Conflicts: Recalibrating Models to Reflect Conflicting Evidencecited by 142
  14. 2023On the Adversarial Robustness of Multi-Modal Foundation Modelscited by 142
  15. 2021A Token-level Reference-free Hallucination Detection Benchmark for Free-form Text Generationcited by 137
  16. 2021VALSE: A Task-Independent Benchmark for Vision and Language Models Centered on Linguistic Phenomenacited by 134
  17. 2023Open Sesame! Universal Black Box Jailbreaking of Large Language Modelscited by 128
  18. 2024Robust CLIP: Unsupervised Adversarial Fine-Tuning of Vision Embeddings for Robust Large Vision-Language Modelscited by 119
  19. 2024Dissecting Adversarial Robustness of Multimodal LM Agentscited by 118
  20. 2022CommonsenseQA 2.0: Exposing the Limits of AI through Gamificationcited by 108
  21. 2021Robustness Gym: Unifying the NLP Evaluation Landscapecited by 107
  22. 2021Measure and Improve Robustness in NLP Models: A Surveycited by 107
  23. 2023The Earth is Flat because...: Investigating LLMs' Belief towards Misinformation via Persuasive Conversationcited by 107
  24. 2021An Empirical Survey of Data Augmentation for Limited Data Learning in NLPcited by 101
  25. 2024The Art of Saying No: Contextual Noncompliance in Language Modelscited by 100
  26. 2022Why Should Adversarial Perturbations be Imperceptible? Rethink the Research Paradigm in Adversarial NLPcited by 99
  27. 2022State-of-the-art generalisation research in NLP: A taxonomy and reviewcited by 97
  28. 2021TopiOCQA: Open-domain Conversational Question Answering with Topic Switchingcited by 95
  29. 2021Competency Problems: On Finding and Removing Artifacts in Language Datacited by 91
  30. 2021QA Dataset Explosion: A Taxonomy of NLP Resources for Question Answering and Reading Comprehensioncited by 91
  31. 2023A Survey of Safety and Trustworthiness of Large Language Models through the Lens of Verification and Validationcited by 91
  32. 2023On the Robustness of Large Multimodal Models Against Image Adversarial Attackscited by 89
  33. 2024Enhancing Noise Robustness of Retrieval-Augmented Language Models with Adaptive Adversarial Trainingcited by 88
  34. 2025Words or Vision: Do Vision-Language Models Have Blind Faith in Text?cited by 86
  35. 2023System 2 Attention (is something you might need too)cited by 85
  36. 2024Fast Adversarial Attacks on Language Models In One GPU Minutecited by 85
  37. 2021Towards Improving Adversarial Training of NLP Modelscited by 79
  38. 2021Towards Interpreting and Mitigating Shortcut Learning Behavior of NLU Modelscited by 78

50 most cited of 568 All 568 (JSON)