arXiv — Machine Learning · · 3 min read

Natural Identifiers for Privacy and Data Audits in Large Language Models

Mirrored from arXiv — Machine Learning for archival readability. Support the source by reading on the original site.

Computer Science > Machine Learning

arXiv:2606.24408 (cs)
[Submitted on 23 Jun 2026]

Title:Natural Identifiers for Privacy and Data Audits in Large Language Models

View a PDF of the paper titled Natural Identifiers for Privacy and Data Audits in Large Language Models, by Lorenzo Rossi and 3 other authors
View PDF
Abstract:Assessing the privacy of large language models (LLMs) presents significant challenges. In particular, most existing methods for auditing differential privacy require the insertion of specially crafted canary data during training, making them impractical for auditing already-trained models without costly retraining. Additionally, dataset inference, which audits whether a suspect dataset was used to train a model, is infeasible without access to a private non-member held-out dataset. Yet, such held-out datasets are often unavailable or difficult to construct for real-world cases since they have to be from the same distribution (IID) as the suspect data. These limitations severely hinder the ability to conduct scalable, post-hoc audits. To enable such audits, this work introduces natural identifiers (NIDs) as a novel solution to the above-mentioned challenges. NIDs are structured random strings, such as cryptographic hashes and shortened URLs, naturally occurring in common LLM training datasets. Their format enables the generation of unlimited additional random strings from the same distribution, which can act as alternative canaries for audits and as same-distribution held-out data for dataset inference. Our evaluation highlights that indeed, using NIDs, we can facilitate post-hoc differential privacy auditing without any retraining and enable dataset inference for any suspect dataset containing NIDs without the need for a private non-member held-out dataset.
Comments: Accepted at ICLR 2026
Subjects: Machine Learning (cs.LG)
Cite as: arXiv:2606.24408 [cs.LG]
  (or arXiv:2606.24408v1 [cs.LG] for this version)
  https://doi.org/10.48550/arXiv.2606.24408
arXiv-issued DOI via DataCite (pending registration)

Submission history

From: Bartłomiej Marek [view email]
[v1] Tue, 23 Jun 2026 10:45:11 UTC (290 KB)
Full-text links:

Access Paper:

    View a PDF of the paper titled Natural Identifiers for Privacy and Data Audits in Large Language Models, by Lorenzo Rossi and 3 other authors
  • View PDF
  • TeX Source

Current browse context:

cs.LG
< prev   |   next >
Change to browse by:
cs

References & Citations

Loading...

BibTeX formatted citation

loading...
Data provided by:

Bookmark

BibSonomy Reddit
Bibliographic Tools

Bibliographic and Citation Tools

Bibliographic Explorer Toggle
Bibliographic Explorer (What is the Explorer?)
Connected Papers Toggle
Connected Papers (What is Connected Papers?)
Litmaps Toggle
Litmaps (What is Litmaps?)
scite.ai Toggle
scite Smart Citations (What are Smart Citations?)
Code, Data, Media

Code, Data and Media Associated with this Article

alphaXiv Toggle
alphaXiv (What is alphaXiv?)
Links to Code Toggle
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub Toggle
DagsHub (What is DagsHub?)
GotitPub Toggle
Gotit.pub (What is GotitPub?)
Huggingface Toggle
Hugging Face (What is Huggingface?)
ScienceCast Toggle
ScienceCast (What is ScienceCast?)
Demos

Demos

Replicate Toggle
Replicate (What is Replicate?)
Spaces Toggle
Hugging Face Spaces (What is Spaces?)
Spaces Toggle
TXYZ.AI (What is TXYZ.AI?)
Related Papers

Recommenders and Search Tools

Link to Influence Flower
Influence Flower (What are Influence Flowers?)
Core recommender toggle
CORE Recommender (What is CORE?)
IArxiv recommender toggle
IArxiv Recommender (What is IArxiv?)
About arXivLabs

arXivLabs: experimental projects with community collaborators

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.

Discussion (0)

Sign in to join the discussion. Free account, 30 seconds — email code or GitHub.

Sign in →

No comments yet. Sign in and be the first to say something.

More from arXiv — Machine Learning