# Trans-RAG: Query-Centric Vector Transformation for Secure Cross-Organizational Retrieval

Source: [Trans-RAG: Query-Centric Vector Transformation for Secure Cross-Organizational Retrieval](<https://www.sota2.com/research/paper/69dc78a87d0172140ec78c66>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-04-14T05:00:11.147Z

Authors: Yu Liu, Kun Peng, Wenxiao Zhang, Fangfang Yuan, Cong Cao, Wenxuan Lu, Yanbing Liu

Published: 2026-04-10T17:58:06.000Z

## Abstract

Retrieval Augmented Generation (RAG) systems deployed across organizational boundaries face fundamental tensions between security, accuracy, and efficiency. Current encryption methods expose plaintext during decryption, while federated architectures prevent resource integration and incur substantial overhead. We introduce Trans-RAG, implementing a novel vector space language paradigm where each organization's knowledge exists in a mathematically isolated semantic space. At the core lies vector2Trans, a multi-stage transformation technique that enables queries to dynamically "speak" each organization's vector space "language" through query-centric transformations, eliminating decryption overhead while maintaining native retrieval efficiency. Security evaluations demonstrate near-orthogonal vector spaces with 89.90{\\deg} angular separation and 99.81% isolation rates. Experiments across 8 retrievers, 3 datasets, and 3 LLMs show minimal accuracy degradation (3.5% decrease in nDCG@10) and significant efficiency improvements over homomorphic encryption.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Information Retrieval on NFCorpus (nDCG@10, Overlap%, Spearman)](<https://www.sota2.com/research/sota/information-retrieval-on-nfcorpus-ndcg-at-10-overlap-percent-spearman.md>): Information Retrieval; NFCorpus; nDCG@10: 37.5.

- [Information Retrieval on FiQA (nDCG@10, Overlap%, Spearman)](<https://www.sota2.com/research/sota/information-retrieval-on-fiqa-ndcg-at-10-overlap-percent-spearman.md>): Information Retrieval; FiQA; nDCG@10: 53.5.

- [Information Retrieval on SciFact (nDCG@10, Overlap%, Spearman)](<https://www.sota2.com/research/sota/information-retrieval-on-scifact-ndcg-at-10-overlap-percent-spearman.md>): Information Retrieval; SciFact; nDCG@10: 0.745.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2604.09541>)

[PDF](<https://arxiv.org/pdf/2604.09541.pdf>)