RAPID: Risk of Attribute Prediction-Induced Disclosure in Synthetic Microdata
Matthias Templ, Oscar Thees, Roman Müller · 2026
Read RAPID: Risk of Attribute Prediction-Induced Disclosure in Synthetic Microdata by Matthias Templ, Oscar Thees, Roman Müller (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-09235.txt.
Description
Fully fake record sets are taking over from older hiding tricks, but old identity-leak tests say little about whether a foe can guess private traits. Matthias Templ and two co-authors bring RAPID, a risk test for trait-guessing leaks under a lifelike attack. In it, the foe trains a model only on released fake data, then runs it on real people's near-ID tags.
For number traits, RAPID gives the share of records guessed within a set error band. For label traits, a base-adjusted trust score shows how much surer the foe is than plain rates alone would hint. The capped, maker-free test is shown with cutoff tuning, doubt gauging, and maker tests, giving a handy top bound on guessing risk.
Catalog record
| CREATOR | Matthias Templ, Oscar Thees, Roman Müller |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 76 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602-09235