An Empirical Analysis of Memorization in Fine-tuned Autoregressive Language Models
Fatemehsadat Mireshghallah, Archit Uniyal, Tianhao Wang, David M Evans, Taylor Berg-Kirkpatrick · 2022
Several recent works have shown that large language models present privacy risks through memorization of training data.Little attention, however, has been given to the fine-tuning phase and it is not well understood how memorization risk varies across different fine-tuning methods (such as fine-tuning the full model, the model head, and adapter).This presents increasing concern as the "pre-train and fine-tune" paradigm proliferates.We empirically study memorization of fine-tuning methods using membership inference and extraction attacks, and show that their susceptibility to attacks is very different.We observe that fine-tuning the head of the model has the highest susceptibility to attacks, whereas fine-tuning smaller adapters appears to be less vulnerable to known extraction attacks.