The paper introduces DEPT, a method that trains a single decoder-only large language model to both expand queries and encode documents for retrieval. By preserving document embeddings close to their initial cached values while allowing gradients to flow through the generator, DEPT stabilizes retrieval targets and enables efficient index reuse and online hard‑negative mining. Experiments on the BEIR benchmark with Qwen3‑4B‑Instruct‑2507 and LLaMA‑3.2‑3B‑Instruct show that DEPT outperforms training‑free, independently trained, and staged unified baselines, with ablations confirming the benefits of preservation, whitening, end‑to‑end expansion training, and online negatives.
By Jingyuan Wang, Richong Zhang, Zhijie Nie, Mingxin Li, Yanzhao Zhang
Multi-vector dense retrieval models, such as ColBERT, achieve strong retrieval effectiveness by modelling fine-grained token-level interactions between queries and documents. Methods such as PLAID use centroid-based quantisation of each token's vector to reduce the index size and speed up retrieval while maintaining strong effectiveness.
Lychee‑Reranker is a fast, not‑late‑interaction reranker that decouples query and passage computation while preserving expressive relevance modeling. It uses an encoder–decoder architecture: passages are pre‑encoded with Matryoshka embedding pooling, and the decoder jointly processes system/user instructions and query intent, with cross‑attention capturing fine‑grained relevance. The design yields four benefits—offline passage encoding for efficiency, cross‑attention expressiveness, compact Matryoshka pooling, and adjustable test‑time compute—and is available in Nano, Small, and Large sizes (0.27B, 1B, and 4B activated parameters).
By Xinping Zhao, Jiaxin Xu, Ziqi Dai, Xin Zhang, Huiyao Chen, Shouzheng Huang, Xianhao Xiong, Danyu Tang, Xinshuo Hu, Guohong Fu, Meishan Zhang, Baotian Hu
arXiv:2609.07262v1 Announce Type: cross
Abstract: Late-interaction visual document retrievers preserve fine-grained page evidence by storing many token embeddings per page, but the resulting storage...
By PS Rishi, Rajeev Ranjan Dwivedi, Vinod K Kurmi
As retrieval systems scale, high-quality reranking becomes increasingly important. However, most existing rerankers, whether encoder-based or decoder-based, jointly encode the query and passage, tightly coupling their computation and limiting deployment efficiency as well as flexibility.
The paper presents a method for improving cross‑lingual retrieval in multilingual encoders by adding a query‑only low‑rank adapter and deterministic routing based on language. The adapter, trained against frozen document embeddings, is used for cross‑lingual queries while the original encoder handles same‑language queries, preserving existing performance. Using the SampoTron LoRA adapter with the Nemotron‑3‑Embed‑1B model, the authors report a 20.9% relative gain in nDCG@10 across six English, Finnish, and Swedish directions on a financial benchmark, with all cross‑lingual directions improving and same‑language performance maintained.
By Akshay Jain, Edward Kim
arXiv:2606.22807v3 Announce Type: replace
Abstract: As retrieval systems scale, effective and efficient reranking becomes increasingly important. However, most existing encoder- and decoder-based rer...
By Xinping Zhao, Jiaxin Xu, Ziqi Dai, Xin Zhang, Huiyao Chen, Shouzheng Huang, Xianhao Xiong, Danyu Tang, Xinshuo Hu, Guohong Fu, Meishan Zhang, Baotian Hu
arXiv:2604.05087v4 Announce Type: replace
Abstract: Generative large language models (LLMs) are increasingly used as inference-time components in retrieval pipelines, for tasks such as query rewritin...
By Omri Uzan, Ron Polonsky, Douwe Kiela, Christopher Potts
The paper introduces a memory-augmented retrieval method that leverages token-level evidence instead of relying solely on single-vector embeddings. By performing sparse token matching and aggregating scores with whole-memory similarity, the approach improves retrieval performance across eight benchmarks and various model families without extra training. An optional low-rank query adapter can further boost results when supervised.
By Han Luo, Zhan Zhuang, Xu He, Jujia Zhao, Yu Li, Suqin Yuan, Dong Li, Lei Feng
arXiv:2605. 30120v3 Announce Type: replace-cross Abstract: Multi-vector retrieval (MVR) models, exemplified by ColBERT, have established new benchmarks in retrieval accuracy by preserving fine-grained token-level interactions.
By Lixuan Guo, Yifei Wang, Tiansheng Wen, Aosong Feng, Stefanie Jegelka, Chenyu You
Multi-vector vision-language retrievers enable fine-grained Visual Document Retrieval (VDR) through late interaction, but storing and scoring hundreds of visual patch embeddings per page incurs substantial overhead. Existing training-free methods rely on pruning or merging: pruning degrades sharply under aggressive compression, whereas merging does not explicitly prioritize important regions when forming representatives.
arXiv:2505. 17810v2 Announce Type: replace Abstract: Approximate nearest neighbor (ANN) search is a performance-critical component of many machine learning pipelines, and rigorous benchmarking is essential for assessing the performance of vector indexes for ANN search.
By Elias J\"a\"asaari, Ville Hyv\"onen, Matteo Ceccarello, Teemu Roos, Martin Aum\"uller