发表机构
New York University; Xi’an Jiaotong University（纽约大学; 西安交通大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本研究揭示持续学习中早期表征变化形成的低维子空间（脚手架）会被后续学习优先重用，且这种重用具有历史依赖性，表明早期经验对网络适应未来任务产生持久的几何影响。
AI 中文摘要
表征随着网络学习新任务而持续变化。我们探究早期表征变化是否自然地形成一种几何结构，并持续影响后续学习。我们识别出早期表征漂移的一个低维子空间，称之为脚手架（scaffold），并测试它是否在后续任务中被重用。在四个预训练视觉编码器和两个数据集上，后续表征变化一致地偏向于这个早期定义的子空间，而非匹配的随机备选子空间。这种重用具有历史依赖性：当网络经历不同的早期任务但后续训练输入相同时，每个网络都优先重用由其自身学习历史诱导的脚手架。同样的偏好出现在单个优化器更新中，尽管网络的主要局部响应方向已偏离原始脚手架。最后，将运动约束在脚手架内对新任务获取的减慢程度大于匹配的随机约束，而对旧任务保持的影响则不太一致。总之，这些结果表明，早期经验在神经网络适应未来任务的方式上留下了持久的几何印记。
英文摘要
Representations continually change as a network learns new tasks. We ask whether early representational changes naturally form a geometric structure that continues to shape later learning. We identify a low-dimensional subspace of early representation drift, which we call a scaffold, and test whether it is reused across subsequent tasks. Across four pretrained visual encoders and two datasets, later representational changes consistently favor this early-defined subspace over matched random alternatives. This reuse is history-dependent: when networks experience different early tasks but identical later training inputs, each network preferentially reuses the scaffold induced by its own learning history. The same preference appears in individual optimizer updates, even though the network's dominant local response directions shift away from the original scaffold. Finally, constraining motion within the scaffold slows new-task acquisition more than matched random constraints, while effects on old-task retention are less consistent. In summary, these results suggest that early experience leaves a persistent geometric imprint on how neural networks adapt to future tasks.Code is available at https://github.com/YuantaoDeng/latent-scaffold.
Comments35 pages, 7 figures
