IDfree-PL samples user-specific embedding distributions on-device to personalize cloud language models without explicit user IDs, with theoretical and empirical support for accuracy gains and embedding-attribution resistance.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
Personalized Language Model Learning on Text Data Without User Identifiers
IDfree-PL samples user-specific embedding distributions on-device to personalize cloud language models without explicit user IDs, with theoretical and empirical support for accuracy gains and embedding-attribution resistance.