ethz-spylab/rlhf_trojan_dataset download history

ethz-spylab/rlhf_trojan_dataset is a dataset on the Hugging Face Hub. In the last 30 days it was downloaded 19 times (6 in the last 7 days), and 26,227 times in total. It ranks #388,347 among datasets by monthly downloads.

Dataset for the competition This is the official dataset for the competition "Find the Trojan: Universal Backdoor Detection in Aligned LLMs" hosted at SaTML 2024. The dataset contains two splits: train and test. Participants should use the train split to execute their proposed methods and

Open ethz-spylab/rlhf_trojan_dataset on Hugging Face

Sister project: Paper Pulse, the upvote history of every Hugging Face Daily Paper.