# Accelerating Large-Scale Inference with Anisotropic Vector Quantization

arXiv: https://arxiv.org/abs/1908.10396 · HTML: https://mlsanity.com/paper/1908.10396/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 27 Aug 2019.

## Built on by (most cited first, 9 of 9)

- [The Benchmark Lottery](https://mlsanity.com/paper/2107.07002/) (2021)
- [Leveraging Large Language Models in Conversational Recommender Systems](https://mlsanity.com/paper/2305.07961/) (2023)
- [Transformer-VQ: Linear-Time Transformers via Vector Quantization](https://mlsanity.com/paper/2309.16354/) (2023)
- [KuaiFormer: Transformer-Based Retrieval at Kuaishou](https://mlsanity.com/paper/2411.10057/) (2024)
- [No Train, all Gain: Self-Supervised Gradients Improve Deep Frozen Representations](https://mlsanity.com/paper/2407.10964/) (2024)
- [Faster Maximum Inner Product Search in High Dimensions](https://mlsanity.com/paper/2212.07551/) (2022)
- [Unleashing the Potential of Two-Tower Models: Diffusion-Based Cross-Interaction for Large-Scale Matching](https://mlsanity.com/paper/2502.20687/) (2025)
- [M3TR: Temporal Retrieval Enhanced Multi-Modal Micro-video Popularity Prediction](https://mlsanity.com/paper/2411.15455/) (2024)
- [Towards Multi-Task Multi-Modal Models: A Video Generative Perspective](https://mlsanity.com/paper/2405.16728/) (2024)

## Abstract (arXiv)

Quantization based techniques are the current state-of-the-art for scaling maximum inner product search to massive databases. Traditional approaches to quantization aim to minimize the reconstruction error of the database points. Based on the observation that for a given query, the database points that have the largest inner products are more relevant, we develop a family of anisotropic quantization loss functions. Under natural statistical assumptions, we show that quantization with these loss functions leads to a new variant of vector quantization that more greatly penalizes the parallel component of a datapoint's residual relative to its orthogonal component. The proposed approach achieves state-of-the-art results on the public benchmarks available at urlann-benchmarks.com.