analiza aiqbench
This research addresses a critical flaw in semi-supervised learning where class distribution mismatches cause unknown samples to be misclassified with high confidence. By introducing a "hub-spoke" latent geometry, the method structurally separates known classes from uncertain outliers, significantly improving feature discriminability and pseudo-label quality. The approach demonstrates robust performance gains by mitigating majority-class domination through explicit uncertainty separation in the feature space.
Hub for Outliers, Spokes for Inliers: Uniform Latent Space Construction for Dual-Mismatched Semi-Supervised Learning
arXiv:2610.07610v1 Announce Type: new Abstract: Semi-supervised learning typically assumes that labeled and unlabeled data share an identical class distribution and label space. However, this setting is often violated: unlabeled data may be imbalanced and contain unknown class samples, causing mismatches in both class distribution and label space. Such dual mismatch leads to majority classes dominating the latent space and unknown class samples being overconfidently misclassified, degrading feature discriminability and pseudo-label quality. To address this, we propose a hub-spoke latent geometry, where known classes are uniformly distributed around a central hub and each class forms compact clusters around its prototype, while the hub provides an anchor for a low-evidence region specifically designed for high-uncertainty unknown class samples. Integrated with an evidence-based classifier, this geometry ultimately enhances feature discriminability and uncertainty separation by mitigating majority-class domination through structured feature organization and guiding high-uncertainty unknown class samples toward the hub. Extensive experiments show that our method outperforms state-of-the-art methods, with a maximum improvement of 3.25% across various settings.
Abstract:Semi-supervised learning typically assumes that labeled and unlabeled data share an identical class distribution and label space. However, this setting is often violated: unlabeled data may be imbalanced and contain unknown class samples, causing mismatches in both class distribution and label space. Such dual mismatch leads to majority classes dominating the latent space and unknown class samples being overconfidently misclassified, degrading feature discriminability and pseudo-label quality. To address this, we propose a hub-spoke latent geometry, where known classes are uniformly distributed around a central hub and each class forms compact clusters around its prototype, while the hub provides an anchor for a low-evidence region specifically designed for high-uncertainty unknown class samples. Integrated with an evidence-based classifier, this geometry ultimately enhances feature discriminability and uncertainty separation by mitigating majority-class domination through structured feature organization and guiding high-uncertainty unknown class samples toward the hub. Extensive experiments show that our method outperforms state-of-the-art methods, with a maximum improvement of 3.25% across various settings.
Equal contribution by Li Yuan and Yaxin Hou. Corresponding author: Yuheng Jia. Emails: {yuan-li,yaxin,230259148,yhjia}@seu.this http URL, gaoyb@qlu.this http URL. 18 pages
Machine Learning (cs.LG)
arXiv:2610.07610 [cs.LG]
(or arXiv:2610.07610v1 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2610.07610 arXiv-issued DOI via DataCite (pending registration)
Text integral sub licența deschisă a publicației · 199 cuvinte
Subiecte conexe
Challenge: Hand coding weights for efficient sequence memorisation
AI Alignment Forum · 23 iul. 2026
Productionizing TurboQuant on AMD GPUs for KV-Cache-Bound LLM Inference
AMD · 11 iun. 2026
I Did This Without Your Knowledge
LessWrong · 17 oct. 2026
DLWM: Diverse Latent World Models for Efficient Multimodal Reasoning
arXiv · 16 iun. 2026