Dense retrieval in multilingual settings often searches over mixed-language collections, yet multilingual embeddings encode language identity alongside semantics. This language signal can inflate similarity for same-language pairs and crowd out relevant evidence written in other languages. We propose LANGSAE EDITING, a post-hoc sparse autoencoder trained on pooled embeddings that enables controllable removal of language-identity signal directly in vector space. The method identifies language-associated latent units using cross-language activation statistics, suppresses these units at inference time, and reconstructs embeddings in the original dimensionality, making it compatible with existing vector databases without retraining the base encoder or re-encoding raw text. Experiments across multiple languages show consistent improvements in ranking quality and cross-language coverage, with especially strong gains for script-distinct languages.
@article{arxiv.2601.04768,
title = {LANGSAE EDITING: Improving Multilingual Information Retrieval via Post-hoc Language Identity Removal},
author = {Dongjun Kim and Jeongho Yoon and Chanjun Park and Heuiseok Lim},
journal= {arXiv preprint arXiv:2601.04768},
year = {2026}
}