SOURCE-LINKED INTELLIGENCE

PRQ-KMeans: Projection Residual Quantization for Semantic ID Tokenization

arXiv · AI, language, vision and robotics · article · Aug 25, 2026 · UTC

Semantic identifiers (SIDs) represent entities as hierarchical token sequences for generative retrieval and recommendation. Residual-quantization tokenizers construct these sequences by selecting a codeword at each level and passing a residual to the next. We view this process as progressive commonality removal: each token captures a component shared within its group, while later tokens should model the remaining differences. This view reveals three limitations: a corpus-wide shared component can consume first-level capacity, hard assignment ignores graded similarities to nearby codewords, and

Read original source ↗ Open in workspace

recordType

paper

region

Global