RQ-MoE: Residual Quantization via Mixture of Experts for Efficient Input-Dependent Vector Compression
This work addresses the decoding bottleneck in dynamic vector quantization for high-dimensional embeddings, offering a practical speed-accuracy trade-off for retrieval and compression tasks.
RQ-MoE introduces a mixture-of-experts framework for input-dependent vector quantization that achieves state-of-the-art reconstruction and retrieval performance while providing 6x-14x faster decoding than prior methods.
Vector quantization is a fundamental tool for compressing high-dimensional embeddings, yet existing multi-codebook methods rely on static codebooks that limit expressiveness under heterogeneous data geometry. While recent dynamic quantizers like QINCo adapt codebooks to individual inputs and improve expressiveness, their strict sequential dependencies create decoding bottlenecks. We propose Residual Quantization via Mixture of Experts (RQ-MoE), a framework combining a two-level MoE with dual-stream quantization to enable input-dependent codebook adaptation for efficient vector quantization. RQ-MoE enables dynamic codebook construction and decouples instruction from quantization, facilitating parallel decoding. Theoretically, we show that standard Residual Quantization and QINCo can be recovered as constrained special cases of RQ-MoE, and derive a guideline for setting expert dimensionality in RQ-MoE. Extensive experiments show that RQ-MoE achieves state-of-the-art or on-par performance in reconstruction and retrieval, while providing 6x-14x faster decoding than prior vector quantization methods. The implementation is available at https://github.com/KDEGroup/RQ-MoE.