发表机构
Queen’s University（女王大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
SPeaR通过冻结预训练网络并在阶段边界插入轻量级可学习模块，直接优化中间表示以应对测试时分布偏移，在多个基准上匹配或超越参数更新方法，且保持源域性能。
AI 中文摘要
测试时自适应（TTA）仅利用无标签的测试数据来解决分布偏移问题。现有方法通常通过更新预训练模型的参数来进行自适应，这限制了可自适应的内容以及网络中可发生自适应的位置。我们转而保持预训练网络冻结，并引导其中间表示。我们引入了SPeaR（用于重对齐表示的引导原语），它在阶段边界处插入轻量级可学习模块，并直接从测试流中优化这些模块，既不需要源数据，也不需要监督式预热。每个原语通过一个门控目标进行优化，该目标仅在自适应有益时降低不确定性，同时使用多样性正则化器防止坍缩，并使用多深度锚点来稳定自适应。我们表明，引导早期表示是最有效的策略，并且相同的原语可跨卷积架构和Transformer架构迁移。在CIFAR-10-C、CIFAR-100-C和ImageNet-C上，SPeaR始终匹配或优于自适应参数数量多出数个数量级的方法，在广泛的批量大小下保持鲁棒性，并在持续自适应过程中保持源域性能。
英文摘要
Test-time adaptation (TTA) addresses distribution shift using only unlabeled test data. Existing methods typically adapt pretrained models by updating their parameters, limiting both what is adapted and where adaptation can occur within the network. We instead keep the pretrained network frozen and steer its intermediate representations. We introduce SPeaR (Steering Primitive for Realigning Representations), which inserts lightweight learnable modules at stage boundaries and optimizes them directly from the test stream, requiring neither source data nor supervised warm-up. Each primitive is optimized using a gated objective that reduces uncertainty only when adaptation is beneficial, along with a diversity regularizer to prevent collapse, and a multi-depth anchor to stabilize adaptation. We show that steering early representations is the most effective strategy, and that the same primitive transfers across convolutional and Transformer architectures. Across CIFAR-10-C, CIFAR-100-C, and ImageNet-C, SPeaR consistently matches or outperforms methods that adapt orders of magnitude more parameters, remains robust across a wide range of batch sizes, and preserves source-domain performance during continual adaptation.
