Metraipolis › archive › dc_03dfacfeea47

CoDe-LoRA: Mitigating the Orthogonality Dilemma in Continual Learning of LLMs via Knowledge Consolidation and Decoupling — arXiv:2610.08312

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.08312
License (as archived)
CC0-1.0
Archived
2026-10-07T11:19:37.304Z
SHA-256 of the archived text
b3a2453a5cf55794993e2fa9c6faebf89eba6f9ba856ee1078318cac8e4a0469
Stable URL of this page
https://metraipolis.com/documento/dc_03dfacfeea47

How to cite

CoDe-LoRA: Mitigating the Orthogonality Dilemma in Continual Learning of LLMs via Knowledge Consolidation and Decoupling — arXiv:2610.08312 — arXiv (en), https://arxiv.org/abs/2610.08312 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_03dfacfeea47 (Metraipolis archive dc_03dfacfeea47, sha256 b3a2453a5cf55794993e2fa9c6faebf89eba6f9ba856ee1078318cac8e4a0469)
BibTeX
@misc{dc03dfacfeea47,
  title = {CoDe-LoRA: Mitigating the Orthogonality Dilemma in Continual Learning of LLMs via Knowledge Consolidation and Decoupling — arXiv:2610.08312},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.08312},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_03dfacfeea47; sha256 b3a2453a5cf55794993e2fa9c6faebf89eba6f9ba856ee1078318cac8e4a0469},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.08312. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Continual learning (CL) is essential for Large Language Models (LLMs) to sequentially adapt to evolving tasks. To mitigate catastrophic forgetting, recent advances implement low-rank adaptation with orthogonal projections (e.g., O-LoRA) to isolate task parameters. However, we reveal that such strict geometric constraints trigger an "Orthogonality Dilemma": rigid parameter isolation impedes the transfer and accumulation of shared representations across semantically related tasks. In this work, we propose a new replay-free method, called Consolidation and Decoupling LoRA (CoDe-LoRA), for CL of LLMs. CoDe-LoRA disentangles the learning process into Consolidating Universal Knowledge and Decoupling Task-Specific Knowledge. To achieve this, CoDe-LoRA leverages an adaptive null space projection mechanism and semantic routing to balance knowledge accumulation with task-specific adaptation. Experimental results across four backbones and three CL benchmarks show that CoDe-LoRA achieves the best average accuracy. Our code is available at this https URL .

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Artificial Intelligence (cs.AI) ; Computation and Language (cs.CL); Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2610.08312 — 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 2026-10-07T11:15:24.604Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).