揭示表格模型在上下文学习中的频谱自适应能力
From Tables to Signals: Revealing Spectral Adaptivity in TabPFN
- 通过信号重建视角分析模型频谱响应
- 上下文样本数决定模型频谱容量,无需调参
- 可直接用于无训练图像去噪,潜力巨大
如TabPFN这类任务无关的表格基础模型在表格学习任务中表现出色,但其归纳偏置的来源仍不清晰。本文从信号重建角度研究TabPFN,首次提供其上下文学习行为的频域分析。结果表明,即使未经超参数调优,TabPFN的有效频率容量也优于标准ReLU-MLP。与仅随训练周期演化的MLP不同,TabPFN的频谱容量能直接适应上下文提供的样本数量,这一现象称为频谱自适应。此外,位置编码会调节其频率响应,类似隐式神经表示的经典结果。最后,这些特性使TabPFN实现无训练、无超参数的图像去噪,展现出作为通用隐式模型在信号重建任务中的潜力。本分析为表格基础模型的结构与归纳偏置提供了新见解。
原文摘要 · Abstract (English)
Task-agnostic tabular foundation models such as TabPFN have achieved impressive performance on tabular learning tasks, yet the origins of their inductive biases remain poorly understood. In this work, we study TabPFN through the lens of signal reconstruction and provide the first frequency-based analysis of its in-context learning behavior. We show that TabPFN possesses a broader effective frequency capacity than standard ReLU-MLPs, even without hyperparameter tuning. Moreover, unlike MLPs whose spectra evolve primarily over training epochs, we find that TabPFN's spectral capacity adapts directly to the number of samples provided in-context, a phenomenon we term Spectral Adaptivity. We further demonstrate that positional encoding modulates TabPFN's frequency response, mirroring classical results in implicit neural representations. Finally, we show that these properties enable TabPFN to perform training-free and hyperparameter-free image denoising, illustrating its potential as a task-agnostic implicit model. Our analysis provides new insight into the structure and inductive biases of tabular foundation models and highlights their promise for broader signal reconstruction tasks.
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。