Mitigating Task-Order Sensitivity and Forgetting via Hierarchical Second-Order Consolidation
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2026
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866908807709327360 |
|---|---|
| author | Nag, Protik Raghavan, Krishnan Narayanan, Vignesh |
| author_facet | Nag, Protik Raghavan, Krishnan Narayanan, Vignesh |
| contents | We introduce $\textbf{Hierarchical Taylor Series-based Continual Learning (HTCL)}$, a framework that couples fast local adaptation with conservative, second-order global consolidation to address the high variance introduced by random task ordering. To address task-order effects, HTCL identifies the best intra-group task sequence and integrates the resulting local updates through a Hessian-regularized Taylor expansion, yielding a consolidation step with theoretical guarantees. The approach naturally extends to an $L$-level hierarchy, enabling multiscale knowledge integration in a manner not supported by conventional single-level CL systems. Across a wide range of datasets and replay and regularization baselines, HTCL acts as a model-agnostic consolidation layer that consistently enhances performance, yielding mean accuracy gains of $7\%$ to $25\%$ while reducing the standard deviation of final accuracy by up to $68\%$ across random task permutations. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2602_02568 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Mitigating Task-Order Sensitivity and Forgetting via Hierarchical Second-Order Consolidation Nag, Protik Raghavan, Krishnan Narayanan, Vignesh Machine Learning We introduce $\textbf{Hierarchical Taylor Series-based Continual Learning (HTCL)}$, a framework that couples fast local adaptation with conservative, second-order global consolidation to address the high variance introduced by random task ordering. To address task-order effects, HTCL identifies the best intra-group task sequence and integrates the resulting local updates through a Hessian-regularized Taylor expansion, yielding a consolidation step with theoretical guarantees. The approach naturally extends to an $L$-level hierarchy, enabling multiscale knowledge integration in a manner not supported by conventional single-level CL systems. Across a wide range of datasets and replay and regularization baselines, HTCL acts as a model-agnostic consolidation layer that consistently enhances performance, yielding mean accuracy gains of $7\%$ to $25\%$ while reducing the standard deviation of final accuracy by up to $68\%$ across random task permutations. |
| title | Mitigating Task-Order Sensitivity and Forgetting via Hierarchical Second-Order Consolidation |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2602.02568 |