English

SLQ: Bridging Modalities via Shared Latent Queries for Retrieval with Frozen MLLMs

Computer Vision and Pattern Recognition 2026-05-12 v3

Abstract

Multimodal Large Language Models (MLLMs) possess intrinsic reasoning and world-knowledge capabilities, yet adapting them for dense retrieval remains challenging. Existing approaches rely on invasive parameter updates, such as full fine-tuning and LoRA, which may disrupt the pre-trained semantic space and impair the structured knowledge essential for reasoning. To address this, we propose SLQ, a parameter-efficient tuning framework that adapts MLLMs for retrieval while keeping the backbone entirely frozen. SLQ introduces a small set of Shared Latent Queries that are appended to both text and image tokens, leveraging the model's native causal attention to aggregate multimodal context into a unified embedding space. Furthermore, to better evaluate retrieval beyond superficial pattern matching, we construct KARR-Bench, a benchmark designed for knowledge-aware reasoning retrieval. Extensive experiments show that SLQ outperforms full fine-tuning and LoRA on COCO and Flickr30K, while achieving competitive performance on MMEB and yielding substantial gains on KARR-Bench, validating that preserving the pre-trained representations via non-invasive adaptation is an effective strategy for MLLM-based retrieval. The code is available under: https://github.com/CnFaker/SLQ.

Keywords

Cite

@article{arxiv.2604.13710,
  title  = {SLQ: Bridging Modalities via Shared Latent Queries for Retrieval with Frozen MLLMs},
  author = {Haoran Lou and Ziyan Liu and Chunxiao Fan and Yuexin Wu and Yue Ming and Hao Wu and Kai Zuo and Yibo Chen and Xu Tang},
  journal= {arXiv preprint arXiv:2604.13710},
  year   = {2026}
}

Comments

Accepted to ICML-2026