BRIGHT
Emerging1papers using it
2025first seen
The 'BRIGHT' benchmark is used to evaluate reasoning-intensive tasks in information retrieval, assessing the performance of models like REARANK in their ability to reason and rerank results effectively.