Skip to content
AI Atlas
PaperActive

Detectable Only Where It Is Confounded: What Verified Duplication Counts Say About Membership Evidence in Language Models

arxiv.org/abs/2609.10830

Updated 22 min ago · first seen 11 Sept 2026

paper_01M294FQ6Q0R0G633GGD0FEF3Z

Published
11 Sept 2026
T1 · 22 min ago
arXiv
2609.10830
T1 · 22 min ago
Category
cs.CL
T1 · 22 min ago

Abstract

When a language model finds a sentence unusually cheap to predict, it is tempting to conclude that the sentence was in its training data. Almost every published test of that inference has had to guess which sentences were in the training data, the members, and which were not. This paper removes the guessing. Two model families, OLMo-2 and Pythia, publish their pretraining corpora, and a public index over those corpora returns the exact number of times any sentence appeared in each. Those counts make three questions answerable directly. The answers form a pincer, closing from two sides. At the duplication levels ordinary text actually has, five models from 1B to 13B parameters carry at most a faint trace of their own exposure. We measure that trace with a design that reads the same sentence through two models, which cancels fluency and quality by construction, and it comes to a rank correlation near -0.08, where -1 would be a perfect relation and 0 none. Where the trace does become strong, above roughly a thousand copies, the two corpora agree on which sentences those are, because they are the famous ones, so exposure can no longer be told apart from fame. Two further measurements show how apparent membership signal gets manufactured. A common way to build a non-member is to change one word of a member. The model does prefer the original, but the gap is the same whether the original appeared once or a hundred times, so what the model is rewarding is the author's word choice, not memory. Above a thousand copies the gap grows with model size on the twelve sentences we can test there, at the same boundary where the pincer closes. And swapping the controls for sentences that differ from the members in register moves a detector from 0.83 to 0.94 AUC, on a scale where 0.5 is a coin flip and 1.0 is perfect separation. We release the sentence banks, counts, and code.

Authors 1

Arman Nik Khah

Specification

Official page

Source:arXiv (Atom API + RSS)T1observed 22 min agohigh

Arxiv announce type
new

Source:arXiv (Atom API + RSS)T1observed 22 min agohigh

arXiv id
2609.10830

Source:arXiv (Atom API + RSS)T1observed 22 min agohigh

Categories
cs.CL, cs.CR, cs.LG

Source:arXiv (Atom API + RSS)T1observed 22 min agohigh

PDF

Source:arXiv (Atom API + RSS)T1observed 22 min agohigh

Primary category
cs.CL

Source:arXiv (Atom API + RSS)T1observed 22 min agohigh

Published
11 Sept 2026

Source:arXiv (Atom API + RSS)T1observed 22 min agohigh

Each value shows its source, tier and observation time. Conflicting claims are kept side by side and flagged — never averaged. How AI Atlas records facts →

Provenance

Attributed facts

9

Source tiers

T19

Freshest observation

22 min ago

Conflicts

None