Control Flow Operators in PyTorch
Yidi Wu, Thomas Bohnstingl, et al.
ICML 2025
Vector quantization is a compression technique for vector data. It creates a collection of codewords to represent the entire vector space. Each vector data is then represented by its nearest neighbor codeword, where the distance between them is the compression error. To improve nearest neighbor representation for vector quantization, we propose to apply sorting transformation to vector data such that members within each vector are sorted. It can be shown that among all permutation transformations, the sorting transformation minimizes L2 distance and maximizes similarity measures such as cosine similarity and Pearson correlation for vector data. Through experimental validation, we show that sorting transformation based vector quantization prominently reduces compression errors and improves nearest neighbor retrieval performance.
Yidi Wu, Thomas Bohnstingl, et al.
ICML 2025
Gosia Lazuka, Andreea Simona Anghel, et al.
SC 2024
Kristjan Greenewald, Yuancheng Yu, et al.
NeurIPS 2024
Ben Fei, Jinbai Liu
IEEE Transactions on Neural Networks