GPIoT: Tailoring Small Language Models for IoT Program Synthesis and Development
Leming Shen 1 , Qiang Yang 2 , Xinyu Huang 1 , Zijing Ma 1 , Yuanqing Zheng 1 · 2025
Read GPIoT: Tailoring Small Language Models for IoT Program Synthesis and Development by Leming Shen 1 , Qiang Yang 2 , Xinyu Huang 1 , Zijing Ma 1 , Yuanqing Zheng 1 (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-00686.txt.
Description
Leming Shen, Qiang Yang, Xinyu Huang, Zijing Ma, and Yuanqing Zheng built GPIoT in 2025, a system that writes code for internet-of-things devices using small language models. Large code models lack the special knowledge these devices need. Fetching that knowledge from the cloud raises privacy, cost, and reliability issues.
GPIoT instead fine-tunes small models that can run locally, using data sets built for this domain. The team also made a benchmark to test such code writing. In experiments and user trials, GPIoT beat leading large code models by a wide margin on task accuracy and earned better user satisfaction. Local running keeps data private and avoids network dropouts.
Catalog record
| CREATOR | Leming Shen 1 , Qiang Yang 2 , Xinyu Huang 1 , Zijing Ma 1 , Yuanqing Zheng 1 |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 72 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2503-00686