Metraipolis › archive › dc_70bf3595906f
BUMP: Brief Unexpected Mid-Interpolation Peak — arXiv:2611.08182
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2611.08182
- License (as archived)
- CC0-1.0
- Archived
- 2026-11-06T14:50:53.777Z
- SHA-256 of the archived text
0230601e0c4e441aee14af9e0a2c31f60c716d45e6ffcaf09ed0cdac13b9485f- Stable URL of this page
https://metraipolis.com/documento/dc_70bf3595906f
How to cite
BUMP: Brief Unexpected Mid-Interpolation Peak — arXiv:2611.08182 — arXiv (en), https://arxiv.org/abs/2611.08182 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_70bf3595906f (Metraipolis archive dc_70bf3595906f, sha256 0230601e0c4e441aee14af9e0a2c31f60c716d45e6ffcaf09ed0cdac13b9485f)
BibTeX
@misc{dc70bf3595906f,
title = {BUMP: Brief Unexpected Mid-Interpolation Peak — arXiv:2611.08182},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2611.08182},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_70bf3595906f; sha256 0230601e0c4e441aee14af9e0a2c31f60c716d45e6ffcaf09ed0cdac13b9485f},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.08182. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
We identify an intriguing phenomenon when linearly interpolating between permutation-aligned neural networks: test accuracy briefly improves near the interpolation endpoints before dropping toward the midpoint, which we term BUMP (Brief Unexpected Mid-interpolation Peak). We observe that BUMP appears consistently across datasets, architectures, training regimes, and remains robust under substantial perturbations such as early checkpoints and extreme pruning. Moreover, controlled ablations suggest that it cannot be attributed to trivial regularization effects. While similar near-endpoint gains are visible in prior interpolation plots, they have not been explicitly documented or systematically and thoroughly analyzed. Leveraging this effect, we propose a simple, novel ensembling scheme: weight-prediction mixed ensembling that combines weight mixing with prediction-space mixing. Across our experiments, the resulting ensembling method consistently outperforms standard logit ensembling in test accuracy and yields improved calibration trade-off, highlighting complementary benefits of weight- and prediction-space mixing. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.LG, nuovi lavori del giorno). Categorie arXiv: Machine Learning (cs.LG) ; Machine Learning (stat.ML). Pagina del lavoro: https://arxiv.org/abs/2611.08182 — 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 2026-11-06T14:21:42.644Z dalla pagina https://arxiv.org/list/cs.LG/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).