彩票路由：面向异构数据的自适应子网络

摘要

在剪枝领域，彩票假说指出大型网络包含稀疏子网络（即中奖彩票），这些子网络可独立训练以达到稠密网络的性能。然而现有方法大多假设存在适用于所有输入的单一通用中奖彩票，忽略了现实数据固有的异质性。本研究提出"路由彩票"（RTL）框架，通过自适应剪枝发现多个专用子网络（称为自适应彩票），每个子网络分别适配不同类别、语义簇或环境条件。在多样化数据集和任务中，RTL在平衡准确率与召回率上持续超越单模型及多模型基线，其参数量比独立模型减少高达10倍，并呈现语义对齐特性。此外，我们发现了激进剪枝下出现的"子网络坍缩"现象，提出基于子网络相似度的无标签诊断方法以识别过度稀疏化问题。本研究将剪枝重新定义为模型结构与数据异质性对齐的机制，为构建更具模块化和环境感知能力的深度学习模型开辟了新路径。

English

In pruning, the Lottery Ticket Hypothesis posits that large networks contain sparse subnetworks, or winning tickets, that can be trained in isolation to match the performance of their dense counterparts. However, most existing approaches assume a single universal winning ticket shared across all inputs, ignoring the inherent heterogeneity of real-world data. In this work, we propose Routing the Lottery (RTL), an adaptive pruning framework that discovers multiple specialized subnetworks, called adaptive tickets, each tailored to a class, semantic cluster, or environmental condition. Across diverse datasets and tasks, RTL consistently outperforms single- and multi-model baselines in balanced accuracy and recall, while using up to 10 times fewer parameters than independent models and exhibiting semantically aligned. Furthermore, we identify subnetwork collapse, a performance drop under aggressive pruning, and introduce a subnetwork similarity score that enables label-free diagnosis of oversparsification. Overall, our results recast pruning as a mechanism for aligning model structure with data heterogeneity, paving the way toward more modular and context-aware deep learning.

彩票路由：面向异构数据的自适应子网络

Routing the Lottery: Adaptive Subnetworks for Heterogeneous Data

摘要

Support