E2Rank: Your Text Embedding can Also be an Effective and Efficient Listwise Reranker

TL;DR

This work introduces E^2Rank, a unified framework that uses a single text embedding model to perform both retrieval and high-quality listwise reranking. By interpreting listwise prompts as pseudo relevance feedback queries and scoring with cosine similarity in embedding space, it achieves strong reranking while avoiding the latency of autoregressive LLM rerankers. The training employs a two-stage process: Stage I with contrastive learning to build embedding quality, and Stage II with a multi-task objective combining InfoNCE and RankNet losses to capture full interactions. Empirically, E^2Rank delivers state-of-the-art reranking performance on BEIR and competitive results on BRIGHT, with substantial efficiency gains and positive transfer to embedding tasks on MTEB, illustrating the practicality and effectiveness of a single model for end-to-end search pipelines.

Abstract

Text embedding models serve as a fundamental component in real-world search applications. By mapping queries and documents into a shared embedding space, they deliver competitive retrieval performance with high efficiency. However, their ranking fidelity remains limited compared to dedicated rerankers, especially recent LLM-based listwise rerankers, which capture fine-grained query-document and document-document interactions. In this paper, we propose a simple yet effective unified framework E2Rank, means Efficient Embedding-based Ranking (also means Embedding-to-Rank), which extends a single text embedding model to perform both high-quality retrieval and listwise reranking through continued training under a listwise ranking objective, thereby achieving strong effectiveness with remarkable efficiency. By applying cosine similarity between the query and document embeddings as a unified ranking function, the listwise ranking prompt, which is constructed from the original query and its candidate documents, serves as an enhanced query enriched with signals from the top-K documents, akin to pseudo-relevance feedback (PRF) in traditional retrieval models. This design preserves the efficiency and representational quality of the base embedding model while significantly improving its reranking performance. Empirically, E2Rank achieves state-of-the-art results on the BEIR reranking benchmark and demonstrates competitive performance on the reasoning-intensive BRIGHT benchmark, with very low reranking latency. We also show that the ranking training process improves embedding performance on the MTEB benchmark. Our findings indicate that a single embedding model can effectively unify retrieval and reranking, offering both computational efficiency and competitive ranking accuracy.

Figures (5)

• Figure 1: (a) Overview of $\text{E}^2$Rank. (b) Average reranking performance on the BEIR benchmark, $\text{E}^2$Rank outperforms other baselines. (c) Reranking latency per query on the Covid dataset.
• Figure 2: Trend of NDCG@10 changes with the number of input documents in listwise prompt.
• Figure 3: Score distribution of using listwise (with 20 documents) and non-listwise prompts.
• Figure 4: Dataset distribution for training.
• Figure 5: Accuracy for labeling the datasets.