Attention in Large Language Models Yields Efficient Zero-Shot Re-Rankers

Connections are not prepared for this record yet, or its source metadata has changed. The original record remains available while background snapshots are rebuilt.