Memorization: A Close Look at Books
To what extent can entire books be extracted from LLMs? Using the Llama 3 70B family of models, and the "prefix-prompting" extraction technique, we were able to auto-regressively reconstruct, with a very high level of similarity, one entire book (Alice's Adventures in Wonderland) from just the first 500 tokens. We were also able to obtain high extraction rates on several other books, piece-wise. However, these successes do not extend uniformly to all books. We show that extraction rates of books correlate with book popularity and thus, likely duplication in the training data. We also confirm the undoing of mitigations in the instruction-tuned Llama 3.1, following recent work (Nasr et al., 2025). We further find that this undoing comes from changes to only a tiny fraction of weights concentrated primarily in the lower transformer blocks. Our results provide evidence of the limits of current regurgitation mitigation strategies and introduce a framework for studying how fine-tuning affects the retrieval of verbatim memorization in aligned LLMs.
Code (0)
등록된 구현이 없습니다.
Tasks
MemorizationMethods 이 논문이 사용한 방법론
Similar Papers 제목 키워드 기반
Extractable Memorization From First Principles
Recent work on extractable memorization in LLMs suffers from two contrasting validity problems. Some studies overstate extraction, e.g., relying on sequences too short to distinguish memorization from predictability. Oth…
Positional Fragility in LLMs: How Offset Effects Reshape Our Understanding of Memorization Risks
Large language models are known to memorize parts of their training data, posing risk of copyright violations. To systematically examine this risk, we pretrain language models (1B/3B/8B) from scratch on 83B tokens, mixin…
AttributeMemorizationExtracting memorized pieces of (copyrighted) books from open-weight language models
Plaintiffs and defendants in copyright lawsuits over generative AI often make sweeping, opposing claims about the extent to which large language models (LLMs) have memorized plaintiffs' protected expression. Drawing on a…
MemorizationSpeak, Memory: An Archaeology of Books Known to ChatGPT/GPT-4
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 m…
MemorizationLearning and Memorization
In the machine learning research community, it is generally believed that there is a tension between memorization and generalization. In this work we examine to what extent this tension exists by exploring if it is …
Memorization