Diving into Kronecker Adapters: Component Design Matters
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2026
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866918530801205248 |
|---|---|
| author | Bai, Jiayu Yu, Danchen Liao, Zhenyu Hou, TianQi Zhou, Feng Qiu, Robert C. Ling, Zenan |
| author_facet | Bai, Jiayu Yu, Danchen Liao, Zhenyu Hou, TianQi Zhou, Feng Qiu, Robert C. Ling, Zenan |
| contents | Kronecker adapters have emerged as a promising approach for fine-tuning large-scale models, enabling high-rank updates through tunable component structures. However, existing work largely treats the component structure as a fixed or heuristic design choice, leaving the dimensions and number of Kronecker components underexplored. In this paper, we identify component structure as a key factor governing the capacity of Kronecker adapters. We perform a fine-grained analysis of both the dimensions and number of Kronecker components. In particular, we show that the alignment between Kronecker adapters and full fine-tuning depends on component configurations. Guided by these insights, we propose Component Designed Kronecker Adapters (CDKA). We further provide parameter-budget-aware configuration guidelines and a tailored training stabilization strategy for practical deployment. Experiments across various architectures and modalities demonstrate the effectiveness of CDKA. Code is available at https://github.com/rainstonee/CDKA. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2602_01267 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Diving into Kronecker Adapters: Component Design Matters Bai, Jiayu Yu, Danchen Liao, Zhenyu Hou, TianQi Zhou, Feng Qiu, Robert C. Ling, Zenan Machine Learning Kronecker adapters have emerged as a promising approach for fine-tuning large-scale models, enabling high-rank updates through tunable component structures. However, existing work largely treats the component structure as a fixed or heuristic design choice, leaving the dimensions and number of Kronecker components underexplored. In this paper, we identify component structure as a key factor governing the capacity of Kronecker adapters. We perform a fine-grained analysis of both the dimensions and number of Kronecker components. In particular, we show that the alignment between Kronecker adapters and full fine-tuning depends on component configurations. Guided by these insights, we propose Component Designed Kronecker Adapters (CDKA). We further provide parameter-budget-aware configuration guidelines and a tailored training stabilization strategy for practical deployment. Experiments across various architectures and modalities demonstrate the effectiveness of CDKA. Code is available at https://github.com/rainstonee/CDKA. |
| title | Diving into Kronecker Adapters: Component Design Matters |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2602.01267 |