WildTrace
Collection
A reliable benchmark for evaluating training data detection in realistic, large-scale post-training settings • 6 items • Updated
YAML Metadata Warning:empty or missing yaml metadata in repo card
Check out the documentation for more information.
Oracle-DPO
Oracle-DPO is a target model developed for evaluating training data detection methods under realistic large-scale direct preference optimization (DPO) settings.
The corresponding benchmark dataset is WildTrace-DPO.
Project Homepage: WildTrace
- Downloads last month
- 50
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support