Gumbel Reranking: Differe
Gumbel Reranking: Differentiable End-to-End Reranker Optimization
Gumbel Reranking: Differentiable End-to-End Reranker Optimization
arXiv:2502.11116v1 Announce Type: new
Abstract: RAG systems rely on rerankers to identify relevant documents. However, fine-tuning these models remains challenging due to the scarcity of annotated query-document pairs. Existing distillation-based approaches suffer from training-inference misalignme…