mlsanityData to 14 Oct 2026

Speak, Memory: An Archaeology of Books Known to ChatGPT/GPT-4

arXiv 2305.00118 ↗First posted 28 Apr 2023cs.CLcited by 125

On the map: LLM safety and security → Machine unlearning

Abstract (arXiv)

In this work, we carry out a data archaeology to infer books that are known to ChatGPT and GPT-4 using a name cloze membership inference query. We find that OpenAI models have memorized a wide collection of copyrighted materials, and that the degree of memorization is tied to the frequency with which passages of those books appear on the web. The ability of these models to memorize an unknown set of books complicates assessments of measurement validity for cultural analytics by contaminating test data; we show that models perform much better on memorized books than on non-memorized books for downstream tasks. We argue that this supports a case for open models whose training data is known.

Key facts. Cited by 125 papers in the map (data through 14 Oct 2026). Placed in Machine unlearning, one of 12 directions in LLM safety and security. Builds on 37 of its 71 references that are in the map. First posted on arXiv 28 Apr 2023.

Builds on

37 of 71 references.

References · 37

  1. 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,399
  2. 2019Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformercited by 14,599
  3. 2023LLaMA: Open and Efficient Foundation Language Modelscited by 14,132
  4. 2022OPT: Open Pre-trained Transformer Language Modelscited by 3,512
  5. 2020The Pile: An 800GB Dataset of Diverse Text for Language Modelingcited by 2,411
  6. 2016Membership Inference Attacks against Machine Learning Modelscited by 2,281
  7. 2020Extracting Training Data from Large Language Modelscited by 2,189
  8. 2022BLOOM: A 176B-Parameter Open-Access Multilingual Language Modelcited by 1,984
  9. 2023Pythia: A Suite for Analyzing Large Language Models Across Training and Scalingcited by 1,786
  10. 2018Datasheets for Datasetscited by 1,386
  11. 2015Aligning Books and Movies: Towards Story-like Visual Explanations by Watching Movies and Reading Bookscited by 1,043
  12. 2019Generalization through Memorization: Nearest Neighbor Language Modelscited by 800
25 more
  1. 2022Quantifying Memorization Across Neural Language Modelscited by 763
  2. 2023Extracting Training Data from Diffusion Modelscited by 732
  3. 2021Deduplicating Training Data Makes Language Models Bettercited by 714
  4. 2021Documenting Large Webtext Corpora: A Case Study on the Colossal Clean Crawled Corpuscited by 492
  5. 2022Large Language Models Struggle to Learn Long-Tail Knowledgecited by 458
  6. 2020Dataset Cartography: Mapping and Diagnosing Datasets with Training Dynamicscited by 425
  7. 2022Deduplicating Training Data Mitigates Privacy Risks in Language Modelscited by 314
  8. 2023Can Large Language Models Transform Computational Social Science?cited by 250
  9. 2022Data Contamination: From Memorization to Exploitationcited by 236
  10. 2022Memorization Without Overfitting: Analyzing the Training Dynamics of Large Language Modelscited by 220
  11. 2022Demystifying Prompts in Language Models via Perplexity Estimationcited by 216
  12. 2022Are Large Pre-Trained Language Models Leaking Your Personal Information?cited by 200
  13. 2021Counterfactual Memorization in Neural Language Modelscited by 156
  14. 2020Question and Answer Test-Train Overlap in Open-Domain Question Answering Datasetscited by 138
  15. 2023Foundation Models and Fair Usecited by 116
  16. 2015The Goldilocks Principle: Reading Children's Books with Explicit Memory Representationscited by 108
  17. 2022Language Contamination Helps Explain the Cross-lingual Capabilities of English Pretrained Modelscited by 92
  18. 2021Memorization vs. Generalization: Quantifying Data Leakage in NLP Performance Evaluationcited by 70
  19. 2022Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictionscited by 57
  20. 2022Whose Language Counts as High Quality? Measuring Language Ideologies in Text Data Selectioncited by 53
  21. 2023Data Portraits: Recording Foundation Model Training Datacited by 30
  22. 2022Heroes, Villains, and Victims, and GPT-3: Automated Extraction of Character Roles Without Training Datacited by 27
  23. 2016Who did What: A Large-Scale Person-Centered Cloze Datasetcited by 24
  24. 2020Living Machines: A study of atypical animacycited by 14
  25. 2022The COVID That Wasn't: Counterfactual Journalism Using GPTcited by 4

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2023: 27; 2024: 38; 2025: 38; 2026: 22.

125 papers.

Sort:

By area:

  1. 2023Large Language Models Are Zero-Shot Time Series Forecasterscited by 500
  2. 2023Detecting Pretraining Data from Large Language Modelscited by 387
  3. 2024Do Membership Inference Attacks Work on Large Language Models?cited by 195
  4. 2024Benchmark Data Contamination of Large Language Models: A Surveycited by 147
  5. 2023Stop Uploading Test Data in Plain Text: Practical Strategies for Mitigating Data Contamination by Evaluation Benchmarkscited by 142
  6. 2023BooookScore: A systematic exploration of book-length summarization in the era of LLMscited by 137
  7. 2024Large Language Models Meet NLP: A Surveycited by 103
  8. 2023Task Contamination: Language Models May Not Be Few-Shot Anymorecited by 99
  9. 2023Can ChatGPT Defend its Belief in Truth? Evaluating LLM Reasoning via Debatecited by 94
  10. 2024One Thousand and One Pairs: A "novel" challenge for long-context language modelscited by 80
  11. 2024No "Zero-Shot" Without Exponential Data: Pretraining Concept Frequency Determines Multimodal Model Performancecited by 75
  12. 2024DE-COP: Detecting Copyrighted Content in Language Models Training Datacited by 74
38 more
  1. 2023SILO Language Models: Isolating Legal Risk In a Nonparametric Datastorecited by 71
  2. 2023GPT4GEO: How a Language Model Sees the World's Geographycited by 68
  3. 2023All Languages Matter: On the Multilingual Safety of Large Language Modelscited by 53
  4. 2023Training Data Extraction From Pre-trained Language Models: A Surveycited by 45
  5. 2024Evaluating Copyright Takedown Methods for Language Modelscited by 43
  6. 2023"According to ...": Prompting Language Models Improves Quoting from Pre-Training Datacited by 42
  7. 2023Large Language Models as Tax Attorneys: A Case Study in Legal Capabilities Emergencecited by 42
  8. 2023Data Contamination Through the Lens of Timecited by 40
  9. 2025Recent Advances in Large Langauge Model Benchmarks against Data Contamination: From Static to Dynamic Evaluationcited by 38
  10. 2023SoK: Memorization in General-Purpose Large Language Modelscited by 37
  11. 2023Position: AI Evaluation Should Learn from How We Test Humanscited by 36
  12. 2024Elephants Never Forget: Memorization and Learning of Tabular Data in Large Language Modelscited by 36
  13. 2024Offset Unlearning for Large Language Modelscited by 36
  14. 2025A Survey on Data Contamination for Large Language Modelscited by 35
  15. 2024SHIELD: Evaluation and Defense Strategies for Copyright Compliance in LLM Text Generationcited by 34
  16. 2024NovelQA: Benchmarking Question Answering on Documents Exceeding 200K Tokenscited by 32
  17. 2024Large language models can be zero-shot anomaly detectors for time series?cited by 31
  18. 2023Who is ChatGPT? Benchmarking LLMs' Psychological Portrayal Using PsychoBenchcited by 28
  19. 2023Experimental Narratives: A Comparison of Human Crowdsourced Storytelling and AI Storytellingcited by 27
  20. 2024Does Data Contamination Detection Work (Well) for LLMs? A Survey and Evaluation on Detection Assumptionscited by 26
  21. 2025The Scales of Justitia: A Comprehensive Survey on Safety Evaluation of LLMscited by 26
  22. 2023People Make Better Edits: Measuring the Efficacy of LLM-Generated Counterfactually Augmented Data for Harmful Language Detectioncited by 25
  23. 2024Digger: Detecting Copyright Content Mis-usage in Large Language Model Trainingcited by 25
  24. 2024CS4: Measuring the Creativity of Large Language Models Automatically by Controlling the Number of Story-Writing Constraintscited by 24
  25. 2024AboutMe: Using Self-Descriptions in Webpages to Document the Effects of English Pretraining Data Filterscited by 22
  26. 2024TimeChara: Evaluating Point-in-Time Character Hallucination of Role-Playing Large Language Modelscited by 22
  27. 2024Con-ReCall: Detecting Pre-training Data in LLMs via Contrastive Decodingcited by 20
  28. 2024Elephants Never Forget: Testing Language Models for Memorization of Tabular Datacited by 19
  29. 2023BHASA: A Holistic Southeast Asian Linguistic and Cultural Evaluation Suite for Large Language Modelscited by 18
  30. 2024Sonnet or Not, Bot? Poetry Evaluation for Large Models and Datasetscited by 18
  31. 2024LLMs and Memorization: On Quality and Specificity of Copyright Compliancecited by 16
  32. 2024MIA-Tuner: Adapting Large Language Models as Pre-training Text Detectorcited by 15
  33. 2025The Emperor's New Clothes in Benchmarking? A Rigorous Examination of Mitigation Strategies for LLM Benchmark Data Contaminationcited by 15
  34. 2025Do LLMs Really Forget? Evaluating Unlearning with Knowledge Correlation and Confidence Awarenesscited by 14
  35. 2024REVS: Unlearning Sensitive Information in Language Models via Rank Editing in the Vocabulary Spacecited by 13
  36. 2024Data Mixture Inference: What do BPE Tokenizers Reveal about their Training Data?cited by 13
  37. 2024Detecting Training Data of Large Language Models via Expectation Maximizationcited by 13
  38. 2024A Comprehensive Survey of Contamination Detection Methods in Large Language Modelscited by 12

50 most cited of 125 All 125 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2025Beyond Public Access in LLM Pre-Training Data
  2. 2026From annotation to reasoning: Culture in language models
  3. 2026CultureBot: A Framework for Retrieval-Augmented Exploration of Cultural Heritage Data
  4. 2026Artificial Intelligence-Assisted Digital Inventory of Cultural Heritage & Traditional Knowledge: Case for Indonesian Open Digital Library of Culture
  5. 2023If the Sources Could Talk: Evaluating Large Language Models for Research Assistance in History
  6. 2025Time Travel: A Comprehensive Benchmark to Evaluate LMMs on Historical and Cultural Artifacts
  7. 2025Remembering Unequally: Global and Disciplinary Bias in LLM Reconstruction of Scholarly Coauthor Lists
  8. 2026Extracting books from production language models