Papers › Did the Neurons Read your Book? Document-level Membership Inference for Large Language Models

Did the Neurons Read your Book? Document-level Membership Inference for Large Language Models

23 Oct 2023arXiv:2310.15007archive 2025-07-28

Matthieu Meeus, Shubham Jain, Marek Rei, Yves-Alexandre de Montjoye

With large language models (LLMs) poised to become embedded in our daily lives, questions are starting to be raised about the data they learned from. These questions range from potential bias or misinformation LLMs could retain from their training data to questions of copyright and fair use of human-generated text. However, while these questions emerge, developers of the recent state-of-the-art LLMs become increasingly reluctant to disclose details on their training corpus. We here introduce the task of document-level membership inference for real-world LLMs, i.e. inferring whether the LLM has seen a given document during training or not. First, we propose a procedure for the development and evaluation of document-level membership inference for LLMs by leveraging commonly used data sources for training and the model release date. We then propose a practical, black-box method to predict document-level membership and instantiate it on OpenLLaMA-7B with both books and academic papers. We show our methodology to perform very well, reaching an AUC of 0.856 for books and 0.678 for papers. We then show our approach to outperform the sentence-level membership inference attacks used in the privacy literature for the document-level membership task. We further evaluate whether smaller models might be less sensitive to document-level inference and show OpenLLaMA-3B to be approximately as sensitive as OpenLLaMA-7B to our approach. Finally, we consider two mitigation strategies and find the AUC to slowly decrease when only partial documents are considered but to remain fairly high when the model precision is reduced. Taken together, our results show that accurate document-level membership can be inferred for LLMs, increasing the transparency of technology poised to change our lives.

PaperPDFCodeCode Syntology ran

In Syntology View this paper on Syntology, its page in Syntology's graph. That page adds which of its functions are also present in other papers' code Syntology holds, with dates, and its dated run record as JSON and a README badge. It also lists the repositories linked to the paper, the abstract and the calls for agents.

Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

For agents, on Syntology's MCP service (how to connect):

Code

Syntology 16 code samples harvested from 1 repository linked to this paper: 13 ran (of which 0 constructed an object rather than computing a result; 13 with no instrument failure: 0 honoured, 0 violated, 13 with no contract checked; 0 where Syntology's instrument failed) · 3 unverified. “Where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code; here there are none.

By repository: official repository: 16 samples from 1 repository, 13 ran (13 with no instrument failure; 0 where Syntology's instrument failed). The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

Levels written on 2026-09-24 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed.

computationalprivacy/document-level-membership-inference officialmentioned in papermentioned on GitHubMITSyntology: 16 samples, 13 ran (0 instrument-failed) · 3 unverified report
ceichler/MIA-bias-removal mentioned on GitHubpytorchSyntology: not harvested report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

16 samples harvested: 13 ran (of which 0 constructed an object rather than computing a result; 13 with no instrument failure: 0 honoured, 0 violated, 13 with no contract checked; 0 where Syntology's instrument failed) · 3 unverified. “Where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Syntology's record for this page has not changed since , the first build that kept a record date for it; when this build read Syntology's graph is in the build record. That date is not when the samples ran.

13ran
3unverified

Licence: 0 of the 16 samples are pointer only, meaning Syntology does not serve that copy's text. This page shows no code text for any sample; each one links to its file in the repository.

Harvested from computationalprivacy/document-level-membership-inference. “Ran” means the sample executed on a synthesized input. It does not mean the output is correct, and nothing here reproduces the paper's results. A line beginning “fixture:” under a sample that ran gives the arguments Syntology's fixture called the sample with, each as name:type/rank/element kind (? where none was recorded), and the type of what the fixture's run returned, exactly as the harness recorded them; the arguments and the returned type were recorded by two separate runs of the same fixture. No output value is shown, and a returned type says nothing about whether the output was right. “Level recorded” gives the UTC date when Syntology's graph wrote the sample's verification level, not when the code executed; where the level was written before Syntology's graph kept that date, the row says “level recorded, date not recorded”. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself. “Our draft was wrong” and “the contract check could not run” are failures of Syntology's instrument, not of the code: the sample ran on Syntology's synthesized input, and then the drafted contract did not describe it, or the check itself raised an error (RAISES), lacked a dependency (DEP_MISSING) or ran out of time (TIMEOUT); hover the label for which. A metamorphic tier, where shown, is evidence of the output's shape, not that the sample implements the method.

Each sample ends with its code_sha256. A sample's code_sha256 is Syntology's identity for that exact code. It is printed in full (16 hexadecimal characters for every sample in this build), and get_code(code_sha256="…") takes it exactly as printed. An agent calls it on Syntology's MCP service (how to connect); click an identity to copy that call.

Repository labels, per sample. official repository: The archive marks this repository official for the paper. named in the paper: The archive records that the paper mentions this repository; it is not marked official. community (archive-listed): In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper. found in paper text by Syntology: Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted. community: Not in the archive's code links for this paper; a community repository Syntology harvested. Samples are grouped by repository: a repository marked official first, then by how directly the repository is tied to the paper, then in the archive's order; never by how many ran. Licence labels name the repository's licence as recorded at harvest. “Pointer only” means Syntology does not serve that copy's text, for one of six reasons: no licence file was found; the licence was not identified; the licence is recorded as permissive but that copy's record is not marked cleared; the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar); this paper's copy was not located, so its licence is not recorded; or this paper's copy is cleared, but the stored copy of the same code that Syntology serves is not. Some licences outside that list permit redistribution, such as WTFPL, and GPL-3.0 under its conditions; they are simply not on the list. Hover a licence label for the reason. File links open the file on GitHub at the default branch, which may have changed since the harvest.

A function or class name links to its page when the name occurs in the harvested code of at least 5 papers; the name is a string, not an identity.

computationalprivacy/document-level-membership-inference official repository 16 samples, 13 ran (0 instrument-failed) · 3 unverified
compute_baselines src/compute_baselines.py ran MIT (permissive) · 650014829c70a9f8 · report fixture: signature text:str/0/str, model1:callable/0/?, model2:callable/0/?, tokenizer1:SimpleTokenizer/0/?, tokenizer2:SimpleTokenizer/0/? · returned dict · level recorded 2026-09-24
extract_features in 73 papers src/feature_extraction.py ran MIT (permissive) · 19d90f1525de7b3c · report fixture: signature data_train:list/1/list, data_test:list/1/list, type:str/0/str · returned tuple · level recorded 2026-09-24
filter_on_date src/split_chunks.py ran MIT (permissive) · 534d1ab80c6edb56 · report fixture: signature dataset:MockDataset/0/?, date_name:str/0/str, min_date:int/0/int, max_date:int/0/int · returned MockDataset · level recorded 2026-09-24
get_diff_max_token src/normalization.py ran MIT (permissive) · 7022375f485ac9be · report fixture: signature idd:int/0/int, dataset:Dataset/0/?, probas:dict/0/?, nlls:dict/0/?, ranks:dict/0/?, token_norm:dict/0/? · returned list · level recorded 2026-09-24
get_norm_val src/normalization.py ran MIT (permissive) · 8fc952dbcf49f67d · report fixture: signature token:str/0/str, token_norm_dict:dict/0/? · returned float · level recorded 2026-09-24
get_token_count src/compute_token_freq.py ran MIT (permissive) · 6dee4ea2c9310a31 · report fixture: signature sample:list/1/dict, token_count:NoneType/0/? · returned dict · level recorded 2026-09-24
get_token_freq src/compute_token_freq.py ran MIT (permissive) · 2bf9d952a5b20ff3 · report fixture: signature token_count:dict/0/? · returned dict · level recorded 2026-09-24
normalize_token_loss src/normalization.py ran MIT (permissive) · 4907842083ba956e · report fixture: signature idd:int/0/int, dataset:Dataset/0/?, nlls:dict/0/?, token_norm:dict/0/? · returned list · level recorded 2026-09-24
remove_small_docs src/split_chunks.py ran MIT (permissive) · 10d15d0c9b1eb682 · report fixture: signature dataset:MinimalDataset/0/?, min_tokens:int/0/int · returned MinimalDataset · level recorded 2026-09-24
scale_features src/meta_classifier.py ran MIT (permissive) · f103f8afd8bf8f81 · report fixture: signature X_train:DataFrame/0/?, X_test:DataFrame/0/? · returned tuple · level recorded 2026-09-24
simple_agg_feat_extraction src/feature_extraction.py ran MIT (permissive) · e9631d636a73e644 · report fixture: signature data:list/1/ndarray, percentiles:tuple/1/int · returned DataFrame · level recorded 2026-09-24
split_data in 68 papers src/meta_classifier.py ran MIT (permissive) · d173585d2fb91fd1 · report fixture: signature member_vals:dict/0/?, non_member_vals:dict/0/?, test_size:float/0/float, seed:int/0/int · returned tuple · level recorded 2026-09-24
validate_clf src/meta_classifier.py ran MIT (permissive) · f7625eef253c7afe · report fixture: signature clf:LogisticRegression/0/?, X_train:DataFrame/0/?, y_train:DataFrame/0/?, X_test:DataFrame/0/?, y_test:DataFrame/0/? · returned dict · level recorded 2026-09-24
calculatePerplexity src/compute_baselines.py unverified MIT (permissive) · 1183adedf54bed5a · report
compute_perplexity in 10 papers src/compute_perplexity.py unverified MIT (permissive) · a2c54d95202e2509 · report
histogram_feats src/feature_extraction.py unverified MIT (permissive) · 9b828cec3990ecfe · report

Tasks

MisinformationSentence

Results from the paper archive 2025-07-28

No leaderboard rows for this paper in the archive.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections