EUVIMEDObjevte EUVIMED
Open access · CC BY 4.0

MATRIX AND VECTOR METHODS FOR LANGUAGE-MODEL EFFICIENCY: DISTINGUISHING EMBEDDING COMPRESSION, TOKEN REDUCTION, AND RAG

Pedro Emílio Amador Salomão, Lívio Almeida Amador

Nexus Science Review · 2026-07-27 · 1 (02) · 1-15

DOI: 10.66104/5vd2bb95 ↗

At a glance

Journal article in Nexus Science Review, published online on 2026-07-27. Written by Pedro Emílio Amador Salomão and 1 co-authors. 15 references. The full text is freely available (open access, CC BY 4.0).

AbstraktBibliografické údajeCiteRelated studies

Abstract (original language)

Language-model efficiency is often discussed through concepts that are related but not equivalent: token count, embedding dimensionality, parameter count, key-value cache size, and retrieved-context length. This article presents a structured critical review, supported by document-level traceability, of matrix and vector methods used to reduce storage and inference costs in language models and retrieval-augmented generation (RAG) systems. The analysis shows that matrix factorization, adaptive embeddings, and Matryoshka representations compress parameters or vectors but do not automatically decrease the number of input tokens. Direct token reduction requires sequence-level interventions, including appropriate tokenization, selection, prompt compression, gist tokens, and retrieved-context control. KV-cache compression reduces memory and latency without necessarily changing the tokens submitted to the model. We propose a mathematical framework that separates V (vocabulary size), d (embedding dimension), n (sequence length), k (number of retrieved documents or passages), and nKV (cache positions retained).

Cite

APA

Salomão, P. E. A., & Amador, L. A. (2026). MATRIX AND VECTOR METHODS FOR LANGUAGE-MODEL EFFICIENCY: DISTINGUISHING EMBEDDING COMPRESSION, TOKEN REDUCTION, AND RAG. Nexus Science Review, 1(02), 1-15. https://doi.org/10.66104/5vd2bb95

Vancouver

Salomão PEA, Amador LA. MATRIX AND VECTOR METHODS FOR LANGUAGE-MODEL EFFICIENCY: DISTINGUISHING EMBEDDING COMPRESSION, TOKEN REDUCTION, AND RAG. Nexus Science Review. 2026;1(02):1-15. doi:10.66104/5vd2bb95

BibTeX
@article{Salomo2026,
  author = {Salomão, Pedro Emílio Amador and Amador, Lívio Almeida},
  title = {MATRIX AND VECTOR METHODS FOR LANGUAGE-MODEL EFFICIENCY: DISTINGUISHING EMBEDDING COMPRESSION, TOKEN REDUCTION, AND RAG},
  journal = {Nexus Science Review},
  year = {2026},
  volume = {1},
  number = {02},
  pages = {1--15},
  doi = {10.66104/5vd2bb95}
}
Start for free →