Metraipolis › archive › dc_80174e1faced
Interpretable structural temperature scaling for calibrating graph neural networks and point-cloud segmentation — arXiv:2611.00175
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2611.00175
- License (as archived)
- CC0-1.0
- Archived
- 2026-11-03T18:35:05.812Z
- SHA-256 of the archived text
8919f91b5da5a681178932611dfd559bc7f13f4152bb82796e5d4c977f6c17f0- Stable URL of this page
https://metraipolis.com/documento/dc_80174e1faced
How to cite
Interpretable structural temperature scaling for calibrating graph neural networks and point-cloud segmentation — arXiv:2611.00175 — arXiv (en), https://arxiv.org/abs/2611.00175 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_80174e1faced (Metraipolis archive dc_80174e1faced, sha256 8919f91b5da5a681178932611dfd559bc7f13f4152bb82796e5d4c977f6c17f0)
BibTeX
@misc{dc80174e1faced,
title = {Interpretable structural temperature scaling for calibrating graph neural networks and point-cloud segmentation — arXiv:2611.00175},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2611.00175},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_80174e1faced; sha256 8919f91b5da5a681178932611dfd559bc7f13f4152bb82796e5d4c977f6c17f0},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.00175. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Graph-aware calibrators make graph neural networks (GNNs) more trustworthy by predicting a temperature for each node, usually with auxiliary networks whose decisions are hard to inspect. We propose Interpretable Structural Temperature Scaling (ISTS), a twelve-parameter linear model that sets each node's temperature from named descriptors of its confidence, neighbourhood and closeness to labelled nodes, so every temperature traces back to descriptor contributions. ISTS preserves predictions, and cross-validated regularisation lets the calibration data decide how far the model departs from temperature scaling (TS). Its explanations are faithful at the level of node groups: on test nodes, ISTS softens groups that a global temperature leaves over-confident and sharpens under-confident ones (Spearman correlation 0.67 over 246 groups). On seven graphs with two backbones and eleven baselines, ISTS has the best average rank in negative log-likelihood (NLL) and Brier score and lowers NLL relative to TS by 1.8%, more than twice the gain of an identically fitted entropy-based temperature. It ranked first again on four held-out graphs under a protocol fixed in advance. From about 100 calibration nodes on, ISTS is a cheap and inspectable choice, and on point-cloud segmentation it significantly improves NLL, Brier score and the ranking of points for rejection. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.LG, nuovi lavori del giorno). Categorie arXiv: Machine Learning (cs.LG) ; Computer Vision and Pattern Recognition (cs.CV). Pagina del lavoro: https://arxiv.org/abs/2611.00175 — 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 2026-11-03T18:12:54.403Z dalla pagina https://arxiv.org/list/cs.LG/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).