Improving Out-of-Distribution Generalization of Neural Rerankers with Contextualized Late Interaction

Open in new window