Info-Coevolution: An Efficient Framework for Data Model Coevolution
Ziheng Qin, Hailun Xu, Wei Chee Yew, Qi Jia, Yang Luo, Kanchan Sarkar, Danhui Guan, Kai Wang, Yang You · 2025
Read Info-Coevolution: An Efficient Framework for Data Model Coevolution by Ziheng Qin, Hailun Xu, Wei Chee Yew, Qi Jia, Yang Luo, Kanchan Sarkar, Danhui Guan, Kai Wang, Yang You (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-2506-08070.txt.
Description
Ziheng Qin and co-authors let models and data grow up together. Their 2025 paper asks a pointed question: given today's model and data, does a fresh sample need a label at all? Stock pipelines keep all data and pay full price. Their frame, Info-Coevolution, labels online and web samples only when the model stands to gain, guided by task models and open models, with no pick bias.
On a flagship image set, label and training costs fell by about a third at no accuracy loss. The frame sets its own saving rate, and semi-supervised tricks halve labeling again. Retrieval can also fatten sets with unlabeled open data.
Catalog record
| CREATOR | Ziheng Qin, Hailun Xu, Wei Chee Yew, Qi Jia, Yang Luo, Kanchan Sarkar, Danhui Guan, Kai Wang, Yang You |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 31 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2506-08070