Jointly selecting layers and modules in transformer foundation models yields better probing features than depth-only or final-layer approaches, and...
论文概览
- arXiv ID:
2611.11531 - 发布机构 / 来源:Hugging Face Daily Papers
- 作者:Ambroise Odonnat, Theo Gnassounou, Laetitia Chapel et al.
- 社区热度:1 Upvotes
核心摘要(Abstract / AI Summary)
Jointly selecting layers and modules in transformer foundation models yields better probing features than depth-only or final-layer approaches, and simple fixed strategies recover much of the gain.
