Empirical benchmarks show distribution similarity between adaptation and pretraining data increases practical privacy leakage in DP-adapted LLMs at fixed theoretical guarantees, with LoRA providing strongest protection for OOD cases.
Performance Measures for Neyman–Pearson Classification , volume =
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Benchmarking Empirical Privacy Protection for Adaptations of Large Language Models
Empirical benchmarks show distribution similarity between adaptation and pretraining data increases practical privacy leakage in DP-adapted LLMs at fixed theoretical guarantees, with LoRA providing strongest protection for OOD cases.