mlsanityData to 14 Oct 2026

FictionalQA: A Dataset for Studying Memorization and Knowledge Acquisition

arXiv 2506.05639 ↗First posted 5 Jun 2025cs.CL cs.LGcited by 0

Not on the map. Why?

Abstract (arXiv)

When language models are trained on textual data, they acquire both knowledge about the structure of language as well as knowledge of facts about the world. At inference time, their knowledge of facts can be leveraged to solve interesting problems and perform useful knowledge work for users. It is well known that language models can verbatim memorize long sequences from their training data. However, it is much less well understood how language models memorize facts seen during training. In this work, we propose a new dataset to specifically empower researchers to study the dual processes of fact memorization and verbatim sequence memorization. The dataset consists of synthetically-generated, webtext-like documents about fictional events, as well as question-answer pairs about the events. We conduct training experiments showing how synthetic data about fictional events can be useful for studying different forms of memorization. We also document some challenges in effectively building realistic, fictional synthetic data.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 38 of its 50 references that are in the map. First posted on arXiv 5 Jun 2025.

Builds on

38 of 50 references.

References · 38

  1. 2020Language Models are Few-Shot Learnerscited by 32,419
  2. 2016Overcoming catastrophic forgetting in neural networkscited by 4,787
  3. 2017TriviaQA: A Large Scale Distantly Supervised Challenge Dataset for Reading Comprehensioncited by 2,955
  4. 2016Membership Inference Attacks against Machine Learning Modelscited by 2,281
  5. 2020Extracting Training Data from Large Language Modelscited by 2,189
  6. 2024Gemma 2: Improving Open Language Models at a Practical Sizecited by 1,998
  7. 2023Pythia: A Suite for Analyzing Large Language Models Across Training and Scalingcited by 1,786
  8. 2019Machine Unlearningcited by 1,061
  9. 2021Membership Inference Attacks From First Principlescited by 878
  10. 2019Right for the Wrong Reasons: Diagnosing Syntactic Heuristics in Natural Language Inferencecited by 876
  11. 2018The Secret Sharer: Evaluating and Testing Unintended Memorization in Neural Networkscited by 864
  12. 2021Deduplicating Training Data Makes Language Models Bettercited by 714
26 more
  1. 2019Certified Data Removal from Machine Learning Modelscited by 535
  2. 2024TOFU: A Task of Fictitious Unlearning for LLMscited by 533
  3. 2018ML-Leaks: Model and Data Independent Membership Inference Attacks and Defenses on Machine Learning Modelscited by 498
  4. 2022Large Language Models Struggle to Learn Long-Tail Knowledgecited by 458
  5. 2023The Reversal Curse: LLMs trained on "A is B" fail to learn "B is A"cited by 407
  6. 2020Label-Only Membership Inference Attackscited by 388
  7. 2024Magpie: Alignment Data Synthesis from Scratch by Prompting Aligned LLMs with Nothingcited by 346
  8. 2022Deduplicating Training Data Mitigates Privacy Risks in Language Modelscited by 314
  9. 2019White-box vs Black-box: Bayes Optimal Strategies for Membership Inferencecited by 271
  10. 2023Physics of Language Models: Part 3.1, Knowledge Storage and Extractioncited by 270
  11. 2020Approximate Data Deletion from Machine Learning Modelscited by 255
  12. 2022Are Large Pre-Trained Language Models Leaking Your Personal Information?cited by 200
  13. 2023Rethinking Benchmark and Contamination for Language Models with Rephrased Samplescited by 195
  14. 2021On the Necessity of Auditable Algorithmic Definitions for Machine Unlearningcited by 171
  15. 2023D4: Improving LLM Pretraining via Document De-Duplication and Diversificationcited by 150
  16. 2024Physics of Language Models: Part 3.3, Knowledge Capacity Scaling Lawscited by 144
  17. 2023Physics of Language Models: Part 3.2, Knowledge Manipulationcited by 143
  18. 2022Measuring Forgetting of Memorized Training Examplescited by 118
  19. 2022Preventing Verbatim Memorization in Language Models Gives a False Sense of Privacycited by 100
  20. 2024How Do Large Language Models Acquire Factual Knowledge During Pretraining?cited by 90
  21. 2025How do language models learn facts? Dynamics, curricula and hallucinationscited by 37
  22. 2024GenQA: Generating Millions of Instructions from a Handful of Promptscited by 35
  23. 2023Students Parrot Their Teachers: Membership Inference on Model Distillationcited by 33
  24. 2025Language Models May Verbatim Complete Text They Were Not Explicitly Trained Oncited by 21
  25. 2024LMD3: Language Model Data Density Dependencecited by 7
  26. 2025Democratizing AI: Open-source Scalable LLM Training on GPU-based Supercomputerscited by 7

Built on by

Not cited yet.