Research map / Label-efficient and robust learning / Test-time adaptation
Architecture-Agnostic Test-Time Adaptation via Backprop-Free Embedding Alignment
MA Xiao, Young D. Kwon, Pan Zhou, Dong Ma
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add layer-wise distance-based weighting to covariance alignment (ablation) | PEA ViT-Base · ImageNet-C | accuracy (%): 25.2 -> 52.9 | Improved |
| add layer-wise weighting (ablation) | PEA ViT-Base · CIFAR100-C | accuracy (%): 67.0 -> 68.3 | Improved |
| add EMA of target statistics (ablation) | PEA ViT-Base · CIFAR100-C | accuracy (%): 68.3 -> 75.7 | Improved |
| add EMA of target statistics (ablation) | PEA ViT-Base · ImageNet-C | accuracy (%): 52.9 -> 64.5 | Improved |
| add lightweight augmentation views (ablation) | PEA ViT-Base · ImageNet-C | accuracy (%): 64.5 -> 66.5 | Improved |
| apply covariance alignment to all layers without weighting (ablation) | PEA ViT-Base · ImageNet-C | accuracy (%): 55.5 -> 25.2 | Hurt |
| apply covariance alignment only (no weighting) (ablation) | PEA ViT-Base · CIFAR100-C | accuracy (%): 61.6 -> 67.0 | Improved |
| use entropy minimization (Tent, EATA) on mixed-domain batches | Tent/EATA ViT-Base · CIFAR100-C mixed domain | accuracy (%): 61.6 -> 61.2 | No clear effect |
| use entropy minimization (Tent) on mixed-domain batches | Tent ResNet-50 · CIFAR100-C mixed domain | accuracy (%): 33.5 -> 17.4 | Hurt |
| reduce batch size from 64 to 4 | PEA ViT-Base · CIFAR100-C | accuracy (%): 77.0 -> 70.0 | Hurt |
| use backprop-free PEA instead of backprop-based Tent | PEA ViT-Base · ImageNet-C memory | memory (MB): 6108 -> 887 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- Tent — Tent: Fully Test-Time Adaptation by Entropy Minimization (ICLR 2021) · better on 3 of 3 shared results
- L-TTA — L-TTA: Lightweight Test-Time Adaptation Using a Versatile Stem Layer (NeurIPS 2024) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Backpropagation-Free Test-Time Adaptation via Probabilistic Gaussian Alignment (NeurIPS 2025 · similarity 0.89)
- Test-time Correlation Alignment (ICML 2025 · similarity 0.89)
- Backpropagation-free Network for 3D Test-time Adaptation (CVPR 2024 · similarity 0.89)
- Efficient Test-time Adaptation through Candidate Verification and Divergence Shifts (NeurIPS 2026 · similarity 0.88)
- A Layer Selection Approach to Test Time Adaptation (AAAI 2025 · similarity 0.88)
- ReservoirTTA: Prolonged Test-time Adaptation for Evolving and Recurring Domains (NeurIPS 2025 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
