Paper Detail

From Joint Risk to Decision Invariance: Decision-Preserving Regularization for Exemplar-Free Class-Incremental Learning

Qi Zhu, Ziang Gan, Wanting Zhang, Libao Zhang

arxiv Score 11.5

Published 2026-11-16 · First seen 2026-11-17

Research Track A

Abstract

Exemplar-free continual learning seeks to minimize joint risk over all seen tasks, yet historical data are unavailable for direct optimization. We observe that, for classification, preserving the model's decisions on previously learned classes helps retain old knowledge. This motivates Decision-Preserving Regularization (DPR). Penalizing changes in all pairwise old-class logit margins reduces algebraically to a centered relative-logit penalty. We further apply spectral whitening to remove the classifier's spectral scale bias. Because the global classifier supplies the decision constraints for the next stage, it must remain calibrated across all seen classes despite the drift of stored old-class statistics during representation updates. We therefore fit a closed-form affine map from paired current-task features to transport old-class Gaussian means and covariances into the updated feature space, then calibrate the global classifier using these distributions and current-task features. Across CIFAR-100, Tiny-ImageNet, ImageNet-Subset, and ImageNet-1K under cold-start protocols, comparisons with state-of-the-art methods show that DPR achieves higher accuracy in most settings while remaining highly efficient. Code is available at https://github.com/bnufw/dpr.

Workflow Status

Review status
pending
Role
unreviewed
Read priority
soon
Vote
Not set.
Saved
no
Collections
Not filed yet.
Next action
Not filled yet.

Reading Brief

No structured notes yet. Add `summary_sections`, `why_relevant`, `claim_impact`, or `next_action` in `papers.jsonl` to enrich this view.

Why It Surfaced

No ranking explanation is available yet.

Tags

No tags.

BibTeX

@article{zhu2026joint,
  title = {From Joint Risk to Decision Invariance: Decision-Preserving Regularization for Exemplar-Free Class-Incremental Learning},
  author = {Qi Zhu and Ziang Gan and Wanting Zhang and Libao Zhang},
  year = {2026},
  abstract = {Exemplar-free continual learning seeks to minimize joint risk over all seen tasks, yet historical data are unavailable for direct optimization. We observe that, for classification, preserving the model's decisions on previously learned classes helps retain old knowledge. This motivates Decision-Preserving Regularization (DPR). Penalizing changes in all pairwise old-class logit margins reduces algebraically to a centered relative-logit penalty. We further apply spectral whitening to remove the cl},
  url = {https://arxiv.org/abs/2611.20619},
  keywords = {cs.CV},
  eprint = {2611.20619},
  archiveprefix = {arXiv},
}

Metadata

{}