mlsanityData to 7 Oct 2026

Whispers of Doubt Amidst Echoes of Triumph in NLP Robustness

arXiv 2311.09694 ↗First posted 16 Nov 2023cs.CLcited by 5

Not on the map. Why?

Abstract (arXiv)

Do larger and more performant models resolve NLP's longstanding robustness issues? We investigate this question using over 20 models of different sizes spanning different architectural choices and pretraining objectives. We conduct evaluations using (a) out-of-domain and challenge test sets, (b) behavioral testing with CheckLists, (c) contrast sets, and (d) adversarial inputs. Our analysis reveals that not all out-of-domain tests provide insight into robustness. Evaluating with CheckLists and contrast sets shows significant gaps in model performance; merely scaling models does not make them adequately robust. Finally, we point out that current approaches for adversarial evaluations of models are themselves problematic: they can be easily thwarted, and in their current forms, do not represent a sufficiently deep probe of model robustness. We conclude that not only is the question of robustness in NLP as yet unresolved, but even some of the approaches to measure robustness need to be reassessed.

Key facts. Cited by 5 papers in the map (data through 7 Oct 2026). Builds on 36 of its 60 references that are in the map. First posted on arXiv 16 Nov 2023.

Builds on

36 of 60 references.

References · 36

  1. 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,204
  2. 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,160
  3. 2020Measuring Massive Multitask Language Understandingcited by 7,768
  4. 2014Explaining and Harnessing Adversarial Examplescited by 7,287
  5. 2022Self-Consistency Improves Chain of Thought Reasoning in Language Modelscited by 5,601
  6. 2016SQuAD: 100,000+ Questions for Machine Comprehension of Textcited by 4,115
  7. 2023Sparks of Artificial General Intelligence: Early experiments with GPT-4cited by 2,579
  8. 2019BoolQ: Exploring the Surprising Difficulty of Natural Yes/No Questionscited by 2,140
  9. 2022Challenging BIG-Bench Tasks and Whether Chain-of-Thought Can Solve Themcited by 1,767
  10. 2020Beyond Accuracy: Behavioral Testing of NLP models with CheckListcited by 979
  11. 2022Super-NaturalInstructions: Generalization via Declarative Instructions on 1600+ NLP Taskscited by 858
  12. 2021Cross-Task Generalization via Natural Language Crowdsourcing Instructionscited by 696
24 more
  1. 2019Is BERT Really Robust? A Strong Baseline for Natural Language Attack on Text Classification and Entailmentcited by 663
  2. 2017Adversarial Examples for Evaluating Reading Comprehension Systemscited by 568
  3. 2020Measuring Robustness to Natural Distribution Shifts in Image Classificationcited by 452
  4. 2020Dataset Cartography: Mapping and Diagnosing Datasets with Training Dynamicscited by 423
  5. 2018QuAC : Question Answering in Contextcited by 379
  6. 2018TextBugger: Generating Adversarial Text Against Real-world Applicationscited by 376
  7. 2018Black-box Generation of Adversarial Text Sequences to Evade Deep Learning Classifierscited by 341
  8. 2016NewsQA: A Machine Comprehension Datasetcited by 330
  9. 2020BAE: BERT-based Adversarial Examples for Text Classificationcited by 316
  10. 2018Stress Test Evaluation for Natural Language Inferencecited by 211
  11. 2021What Will it Take to Fix Benchmarking in Natural Language Understanding?cited by 200
  12. 2022WANLI: Worker and AI Collaboration for Natural Language Inference Dataset Creationcited by 200
  13. 2021CrossFit: A Few-shot Learning Challenge for Cross-task Generalization in NLPcited by 142
  14. 2023WiCE: Real-World Entailment for Claims in Wikipediacited by 118
  15. 2021Robustness Gym: Unifying the NLP Evaluation Landscapecited by 107
  16. 2021Measure and Improve Robustness in NLP Models: A Surveycited by 107
  17. 2022State-of-the-art generalisation research in NLP: A taxonomy and reviewcited by 97
  18. 2023Revisiting Out-of-distribution Robustness in NLP: Benchmark, Analysis, and LLMs Evaluationscited by 96
  19. 2020Reevaluating Adversarial Examples in Natural Languagecited by 83
  20. 2022GLUE-X: Evaluating Natural Language Understanding Models from an Out-of-distribution Generalization Perspectivecited by 72
  21. 2020What do Models Learn from Question Answering Datasets?cited by 45
  22. 2021The Dangers of Underclaiming: Reasons for Caution When Reporting How NLP Systems Failcited by 38
  23. 2022CONDAQA: A Contrastive Reading Comprehension Dataset for Reasoning about Negationcited by 38
  24. 2022Residue-Based Natural Language Adversarial Attack Detectioncited by 8

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2024: 2; 2025: 2; 2026: 1.

5 papers.

  1. 2024Beyond Perplexity: Multi-dimensional Safety Evaluation of LLM Compressioncited by 32
  2. 2024Chain-of-Probe: Examining the Necessity and Accuracy of CoT Step-by-Stepcited by 8
  3. 2025Pay Attention to Real World Perturbations! Natural Robustness Evaluation in Machine Reading Comprehensioncited by 2
  4. 2026Measuring Distribution Shift in User Prompts and Its Effects on LLM Performancecited by 1
  5. 2025Do Generalisation Results Generalise?cited by 0