Certifying Whether to Adapt: Baseline-Relative PAC-Bayes-MMD Guarantees for Test-Time Adaptation
Abstract
Test-time adaptation (TTA) needs a decision about an update, not only an absolute-risk estimate. We give PAC-Bayesian certificates for the target excess risk of a candidate relative to a pilot-fixed baseline. A radial certificate uses an MMD ambiguity radius; a direction-sensitive certificate retains the sign of the empirical RKHS shift and can therefore certify target improvement even when source risk increases. For finite binary state spaces we derive statewise marginal and signed conditional-shift refinements, with penalties localized to candidate--baseline disagreements. A source-worsening update with target excess risk −0.576 is accepted in all 1,000 repetitions by the directional certificate and never by the radial certificate; learned finite-state updates with excess risks −0.800 and −0.401 are likewise accepted in every repetition. On Covertype, the certificate safely rejects all updates and blocks geographic-shift adaptations whose unconditional deployment increases risk by up to $2.75$ percentage points in prespecified base libraries; an exploratory stress test reaches $5.66$ points. A feature-quantized CIFAR-10-C and ImageNet-C study is structurally vacuous and is reported only as a diagnostic limitation, not as a practical neural certificate.
Related
- Certifying Distributionally Robust Decisions from Projection Observations
- Perfect Reconstruction Cannot Show Whether Two Codebooks on a VAE's Latent Space Agree
- Certifying non-existence of integer points on curves
- Stand-In Effects and Pooling Amplification: Design-Based Tests of Whether Biological World Models Use Actions
- Spread and Scale: What Determines Whether Test-Time Budget Allocation Pays
BibTeX
@misc{ariq2026certifying,
title = {{Certifying Whether to Adapt: Baseline-Relative PAC-Bayes-MMD Guarantees for Test-Time Adaptation}},
author = {Ahanaf Hasan Ariq and Hasaan Ahmad and Aayush Parekh and Abdul Monaf Chowdhury},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2605.21783},
eprint = {2605.21783},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2605.21783},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Certifying Whether to Adapt: Baseline-Relative PAC-Bayes-MMD Guarantees for Test-Time Adaptation
AU - Ahanaf Hasan Ariq
AU - Hasaan Ahmad
AU - Aayush Parekh
AU - Abdul Monaf Chowdhury
PY - 2026
DA - 2026/10/18
PB - arXiv
DO - 10.48550/arxiv.2605.21783
UR - https://arxiv.org/abs/2605.21783
AB - Test-time adaptation (TTA) needs a decision about an update, not only an absolute-risk estimate. We give PAC-Bayesian certificates for the target excess risk of a candidate relative to a pilot-fixed baseline. A radial certificate uses an MMD ambiguity radius; a direction-sensitive certificate retains the sign of the empirical RKHS shift and can therefore certify target improvement even when source risk increases. For finite binary state spaces we derive statewise marginal and signed conditional-shift refinements, with penalties localized to candidate--baseline disagreements. A source-worsening update with target excess risk −0.576 is accepted in all 1,000 repetitions by the directional certificate and never by the radial certificate; learned finite-state updates with excess risks −0.800 and −0.401 are likewise accepted in every repetition. On Covertype, the certificate safely rejects all updates and blocks geographic-shift adaptations whose unconditional deployment increases risk by up to $2.75$ percentage points in prespecified base libraries; an exploratory stress test reaches $5.66$ points. A feature-quantized CIFAR-10-C and ImageNet-C study is structurally vacuous and is reported only as a diagnostic limitation, not as a practical neural certificate.
ER -