When Is 0.1% Enough? Analyzing the Combined Effects of Dimensionality Reduction and Quantization on Text Embedding Compression
Fuente:
arXiv
Salvato in:
| Autori principali: | , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2026
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866913176881201152 |
|---|---|
| author | Kisako, Riku Tsukagoshi, Hayato Sasano, Ryohei |
| author_facet | Kisako, Riku Tsukagoshi, Hayato Sasano, Ryohei |
| contents | Recent high-performing text embedding models often output high-dimensional real-valued vectors, resulting in substantial storage and computational costs. To address this issue, compression methods based on dimensionality reduction or quantization have been proposed; however, the effects of combining dimensionality reduction and quantization have not been sufficiently investigated. In this paper, we systematically examine the effectiveness of compressing text embeddings by combining dimensionality reduction and quantization, using four MTEB task families and four pretrained embedding models. The experimental results demonstrate that combining dimensionality reduction and quantization enables substantially stronger compression than using either method alone, that in some settings embeddings can be reduced to as little as 0.1% of their original size with almost no performance degradation, and that the optimal compression strategy depends on the task. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2606_01074 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | When Is 0.1% Enough? Analyzing the Combined Effects of Dimensionality Reduction and Quantization on Text Embedding Compression Kisako, Riku Tsukagoshi, Hayato Sasano, Ryohei Computation and Language Recent high-performing text embedding models often output high-dimensional real-valued vectors, resulting in substantial storage and computational costs. To address this issue, compression methods based on dimensionality reduction or quantization have been proposed; however, the effects of combining dimensionality reduction and quantization have not been sufficiently investigated. In this paper, we systematically examine the effectiveness of compressing text embeddings by combining dimensionality reduction and quantization, using four MTEB task families and four pretrained embedding models. The experimental results demonstrate that combining dimensionality reduction and quantization enables substantially stronger compression than using either method alone, that in some settings embeddings can be reduced to as little as 0.1% of their original size with almost no performance degradation, and that the optimal compression strategy depends on the task. |
| title | When Is 0.1% Enough? Analyzing the Combined Effects of Dimensionality Reduction and Quantization on Text Embedding Compression |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2606.01074 |