Metraipolis › archive › dc_ebd83d6f568a
BaCP: Backbone Contrastive Pruning for Preserving Representations in Extremely Sparse Neural Networks — arXiv:2610.02524
- Source
- arXiv
- Original
- https://arxiv.org/abs/2610.02524
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-05T23:09:29.920Z
- SHA-256 of the archived text
22968ea509ae676a6fe2ad193718973f8d494b5dcb64abcea82164b75ccf84cb- Stable URL of this page
https://metraipolis.com/documento/dc_ebd83d6f568a
How to cite
BaCP: Backbone Contrastive Pruning for Preserving Representations in Extremely Sparse Neural Networks — arXiv:2610.02524 — arXiv, https://arxiv.org/abs/2610.02524 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_ebd83d6f568a (Metraipolis archive dc_ebd83d6f568a, sha256 22968ea509ae676a6fe2ad193718973f8d494b5dcb64abcea82164b75ccf84cb)
BibTeX
@misc{dcebd83d6f568a,
title = {BaCP: Backbone Contrastive Pruning for Preserving Representations in Extremely Sparse Neural Networks — arXiv:2610.02524},
howpublished = {arXiv},
url = {https://arxiv.org/abs/2610.02524},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_ebd83d6f568a; sha256 22968ea509ae676a6fe2ad193718973f8d494b5dcb64abcea82164b75ccf84cb},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv, https://arxiv.org/abs/2610.02524. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Unstructured pruning at extreme sparsity often suffers from representational collapse, causing sharp drops in accuracy. To address this, we study Backbone Contrastive Pruning (BaCP), which regularizes the sparse network's embedding space by aligning it with pretrained, fine-tuned, and historical snapshot models. Building on the contrastive decomposition of the CAP framework (Xu et al., 2022), we provide a rigorous matched-budget characterization of this approach across multiple pruning criteria. Evaluated across 90 settings, BaCP improves accuracy substantially in extreme sparsity regimes where standard pruning fails, and is close to baseline where representations remain intact. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.LG, nuovi lavori del giorno). Categorie arXiv: Machine Learning (cs.LG). Pagina del lavoro: https://arxiv.org/abs/2610.02524 — 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 2026-10-05T23:01:38.597Z dalla pagina https://arxiv.org/list/cs.LG/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).