Construct · 1 fields · 1 citations

training data extraction attack: a standalone construct

One name. One construct. A landmark on the map.

● anchor · a standalone construct Nearest construct by meaning·not indexed

A single standalone construct; nearby constructs are located against it in the Collider. Tap a field for its definition, unit, and measures.

Adversarial machine learning

querying a language model to recover verbatim training sequences, including personal information, from its outputs.

In adversarial machine learning, training data extraction attack refers to querying a language model to recover verbatim training sequences, including personal information, from its outputs. This meaning comes from Carlini (2020) in arXiv.[1]

Unit
Recovery means
Measured by

Run it through the Collider, or take the citations with you.

See how training data extraction attack sits against neighbouring constructs in the Collider, or export every reference on this page in one click.

Open in Collider Export BibTeX Report this entry

Nearby constructs

attack4 fields · 10 citationsjingle · one word, four constructs information4 fields · 5 citationsjingle · one word, four constructs learning6 fields · 14 citationsjingle · one name, six constructs

References

Carlini, N., Tramer, F., Wallace, E., Jagielski, M., Herbert-Voss, A., Lee, K., Roberts, A., Brown, T., Song, D., Erlingsson, U., Oprea, A., & Raffel, C. (2020). Extracting Training Data from Large Language Models. arXiv. https://doi.org/10.48550/arxiv.2012.07805

Cited on this page?

Confirm your construct, or correct us.

If your work is cited behind one of these entries, you know better than we do whether we captured the meaning right. Confirm it, suggest an edit, or dispute it, and your name goes on the entry as an author who has checked it, linked to your ORCID.

Claim your construct