不更新模型参数，通过调整隐空间分布实现少样本自适应。
Rethinking Test-Time Training: Tilting The Latent Distribution For Few-Shot Source-Free Adaptation
- 用小样本支持集计算任务相似度，对编码器输出的隐空间分布进行指数倾斜重加权。
- 在多个基准上性能接近甚至超过需调参的方法，且全程冻结模型。
- 适合部署时无法更新模型、无上游数据的场景，如边缘设备或隐私敏感应用。
在部署环境中，即使轻量级参数更新（如参数高效微调）也可能引发模型偏移或训练不稳定。本文研究在完全冻结模型、且无上游数据可用的前提下，对基础模型进行少样本分类的测试时自适应。提出迄今首个无需训练的推理方法，通过改变编码器生成的隐空间分布来调整预测结果。利用小规模带标签支持集计算任务相似度得分，采用指数倾斜方式在KL最优意义下重新加权隐空间分布，不修改任何模型参数。实验表明，该方法在多个基准和不同样本量设置下均能持续媲美基于参数更新的方法，同时满足更严格普遍的约束条件。结果证明，在全冻结模型架构下，仅通过推理阶段的分布修正即可实现有效的测试时自适应。
原文摘要 · Abstract (English)
Often, constraints arise in deployment settings where even lightweight parameter updates e.g. parameter-efficient fine-tuning could induce model shift or tuning instability. We study test-time adaptation of foundation models for few-shot classification under a completely frozen-model regime, where additionally, no upstream data are accessible. We propose arguably the first training-free inference method that adapts predictions to the new task by performing a change of measure over the latent embedding distribution induced by the encoder. Using task-similarity scores derived from a small labeled support set, exponential tilting reweights latent distributions in a KL-optimal manner without modifying model parameters. Empirically, the method consistently competes with parameter-update-based methods across multiple benchmarks and shot regimes, while operating under strictly and universally stronger constraints. These results demonstrate the viability of inference-level distributional correction for test-time adaptation even with a fully-frozen model pipeline.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。