Fine-tuning can Help Detect Pretraining Data from Large Language Models.

Hengxiang Zhang, Songxin Zhang, Bingyi Jing, Hongxin Wei

Browse the full ICLR paper archive.