Metraipolis › archive › dc_54ec024163b0

Dynamic LoRA-Experts and Prototype-Ensemble Matching for Class-Incremental Learning — arXiv:2609.39839

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.39839
License (as archived)
CC0-1.0
Archived
2026-10-07T13:46:43.796Z
SHA-256 of the archived text
dce7eeefa2eaa4ff0e015c3daf3264fd55193b0c7e76657cccaed86bfffbf125
Stable URL of this page
https://metraipolis.com/documento/dc_54ec024163b0

How to cite

Dynamic LoRA-Experts and Prototype-Ensemble Matching for Class-Incremental Learning — arXiv:2609.39839 — arXiv (en), https://arxiv.org/abs/2609.39839 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_54ec024163b0 (Metraipolis archive dc_54ec024163b0, sha256 dce7eeefa2eaa4ff0e015c3daf3264fd55193b0c7e76657cccaed86bfffbf125)
BibTeX
@misc{dc54ec024163b0,
  title = {Dynamic LoRA-Experts and Prototype-Ensemble Matching for Class-Incremental Learning — arXiv:2609.39839},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.39839},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_54ec024163b0; sha256 dce7eeefa2eaa4ff0e015c3daf3264fd55193b0c7e76657cccaed86bfffbf125},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.39839. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Class-Incremental Learning (CIL) aims to continuously learn new classes without forgetting previously acquired knowledge. Parameter-efficient fine-tuning with pre-trained models reduces parameter overhead but can suffer from cumulative interference and suboptimal alignment between inference samples and specialized modules. We propose Dynamic LoRA-Experts and Prototype-Ensemble Matching (DLEPEM), a two-stage rehearsal-free framework. DLEPEM allocates a task-specific LoRA-Expert for each incremental task to reduce cross-task interference, then combines frozen pre-trained-model prototypes with task-adaptive LoRA-Expert prototypes for reliable task-level discrimination. Experiments on standard CIL and Few-Shot CIL benchmarks demonstrate strong performance under the evaluated protocols.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-10-01 in cs.LG).
Categorie arXiv: Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2609.39839 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-07T13:18:10.474Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-10-01?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).