Reliability-Gated Source Anchoring for Continual Test-Time Adaptation
Vikash Singh, Debargha Ganguly, Weicong Chen, Sabyasachi Sahoo, Sreehari Sankar, Biyao Zhang, Mohsen Hariri, Shouren Wang, Osama Zafar, Christian Gagn\'e, Vipin Chaudhary

TL;DR
This paper introduces RMemSafe, a reliability-gated method for continual test-time adaptation that dynamically reduces reliance on source anchors when source reliability diminishes, improving robustness on corrupted data.
Contribution
It proposes a novel entropy-based gating mechanism for source anchoring in CTTA, enhancing adaptation robustness under source degradation.
Findings
RMemSafe achieves lowest error on 8 of 9 continual-corruption tasks.
It improves ROID+ASR by 1.05 percentage points on ResNet-50.
The entropy gate effectively detects source collapse, enabling graceful adaptation.
Abstract
Continual test-time adaptation (CTTA) updates a pretrained model online on an unlabeled, non-stationary stream while anchoring it to a frozen source checkpoint. This anchor is useful only when the source remains reliable. On CCC-Hard, however, a ResNet-50 source falls to approximately top- accuracy, while existing source-anchored CTTA methods continue applying the same anchor strength. We call this failure mode blind anchoring and propose RMemSafe, a reliability-gated extension of ROID that uses the frozen source's normalized predictive entropy to attenuate all explicit source-coupled uses in the objective. When the source posterior approaches uniformity, the gate closes: the source anchor and agreement filter vanish, and the objective reduces to a source-agnostic fallback comprising ROID's base losses plus marginal calibration. Combined with ASR, RMemSafe achieves the lowest…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
