Enter a query and a list of candidate documents (one per line). The model ranks them by relevance using listwise causal attention.
Model card · Blog
Ranked Results