Guess or Recall? Training CNNs to Classify and Localize Memorization in LLMs

Jérémie Dentan, Davide Buscaldi, Sonia Vanier · 2025

Read Guess or Recall? Training CNNs to Classify and Localize Memorization in LLMs by Jérémie Dentan, Davide Buscaldi, Sonia Vanier (2025) 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-2508-02573.txt.

Description

Jérémie Dentan, Davide Buscaldi, and Sonia Vanier probe how language models memorize text word for word, in a paper from twenty twenty-five. A model may repeat a passage because it truly stored it, or because the passage is easy to guess. Sorting these cases matters for privacy and for science.

The team trained image-style networks on the model's internal attention maps to sort samples by type. The old sorting scheme matched those maps poorly. Their new scheme has three kinds: guessed samples, samples recalled after heavy repetition in training, and non-memorized ones. Much of what looks extractable turns out to be mere guessing. A custom visual tool pinpoints which attention regions serve each kind.

Catalog record

CREATORJérémie Dentan, Davide Buscaldi, Sonia Vanier
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE62 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2508-02573

Browse the full library · About the Tower of Babel