AcTTA: Rethinking Test-Time Adaptation via Dynamic Activation

Abstract

Test-time adaptation (TTA) aims to mitigate performance degradation under distribution shifts by updating model parameters during inference. Existing approaches have primarily framed adaptation around affine modulation, focusing on recalibrating normalization layers. This perspective, while effective, overlooks another influential component in representation dynamics: the activation function. We revisit this overlooked space and propose AcTTA, an activation-aware framework that reinterprets conventional activation functions from a learnable perspective and updates them adaptively at test time. AcTTA reformulates conventional activation functions (e.g., ReLU, GELU) into parameterized forms that shift their response threshold and modulate gradient sensitivity, enabling the network to adjust activation behavior under domain shifts. This functional reparameterization enables continuous adjustment of activation behavior without modifying network weights or requiring source data. Despite its simplicity, AcTTA achieves robust and stable adaptation across diverse corruptions. Across CIFAR10-C, CIFAR100-C, and ImageNet-C, AcTTA consistently surpasses normalization-based TTA methods. Our findings highlight activation adaptation as a compact and effective route toward domain-shift-robust test-time learning, broadening the prevailing affine-centric view of adaptation.

Figures (3)

• Figure 1: Visualization of the adaptive activation function. (a) The base activation functions $\phi(x)$. (b) The AcTTA activation output $g(x)$; each parameter set $[\,\cdot,\cdot,\cdot\,]$ denotes $[\lambda^{+},\,\lambda^{-},\,c]$. (c) The first derivative $g'(x)$ with varying $(\lambda^{+},\lambda^{-})$ (d) The effect of the shift parameter $c$, which avoids strict zero-centering during adaptation.
• Figure 2: Visualizations of adapted activations during AcTTA. AcTTA-ReLU results are obtained from WRN-28 on CIFAR10-C after completing adaptation on the Gaussian noise corruption. AcTTA-GELU results are from ViT-B/16 on ImageNet-C at the same adaptation endpoint. The learned activation exhibits flexible, channel-specific shaping behavior, adapting its form according to the input distribution. The three legend values denote the adapted parameter values of AcTTA, $[\lambda^{+},\,\lambda^{-},\,c]$, corresponding to positive/negative slope modulation and shift.
• Figure 3: Gradient pass-through ratio.