LLMComp: A Language Modeling Paradigm for Error-Bounded Scientific Data Compression (Technical Report)
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866912688134684672 |
|---|---|
| author | Li, Guozhong Alhumaidi, Muhannad Skiadopoulos, Spiros Kalnis, Panos |
| author_facet | Li, Guozhong Alhumaidi, Muhannad Skiadopoulos, Spiros Kalnis, Panos |
| contents | The rapid growth of high-resolution scientific simulations and observation systems is generating massive spatiotemporal datasets, making efficient, error-bounded compression increasingly important. Meanwhile, decoder-only large language models (LLMs) have demonstrated remarkable capabilities in modeling complex sequential data. In this paper, we propose LLMCOMP, a novel lossy compression paradigm that leverages decoder-only large LLMs to model scientific data. LLMCOMP first quantizes 3D fields into discrete tokens, arranges them via Z-order curves to preserve locality, and applies coverage-guided sampling to enhance training efficiency. An autoregressive transformer is then trained with spatial-temporal embeddings to model token transitions. During compression, the model performs top-k prediction, storing only rank indices and fallback corrections to ensure strict error bounds. Experiments on multiple reanalysis datasets show that LLMCOMP consistently outperforms state-of-the-art compressors, achieving up to 30% higher compression ratios under strict error bounds. These results highlight the potential of LLMs as general-purpose compressors for high-fidelity scientific data. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2510_23632 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | LLMComp: A Language Modeling Paradigm for Error-Bounded Scientific Data Compression (Technical Report) Li, Guozhong Alhumaidi, Muhannad Skiadopoulos, Spiros Kalnis, Panos Machine Learning Artificial Intelligence The rapid growth of high-resolution scientific simulations and observation systems is generating massive spatiotemporal datasets, making efficient, error-bounded compression increasingly important. Meanwhile, decoder-only large language models (LLMs) have demonstrated remarkable capabilities in modeling complex sequential data. In this paper, we propose LLMCOMP, a novel lossy compression paradigm that leverages decoder-only large LLMs to model scientific data. LLMCOMP first quantizes 3D fields into discrete tokens, arranges them via Z-order curves to preserve locality, and applies coverage-guided sampling to enhance training efficiency. An autoregressive transformer is then trained with spatial-temporal embeddings to model token transitions. During compression, the model performs top-k prediction, storing only rank indices and fallback corrections to ensure strict error bounds. Experiments on multiple reanalysis datasets show that LLMCOMP consistently outperforms state-of-the-art compressors, achieving up to 30% higher compression ratios under strict error bounds. These results highlight the potential of LLMs as general-purpose compressors for high-fidelity scientific data. |
| title | LLMComp: A Language Modeling Paradigm for Error-Bounded Scientific Data Compression (Technical Report) |
| topic | Machine Learning Artificial Intelligence |
| url | https://arxiv.org/abs/2510.23632 |