Hyperspherical Representation Learning

Ask about this topic

Overview

Hyperspherical representation learning studies how to encode data as directions on a unit sphere, where angles express similarity and embeddings have controlled magnitude. Common approaches include cosine-based contrastive learning, class prototypes, and probability distributions such as von Mises–Fisher, Power Spherical, and Kent models, alongside regularizers that shape the spread of embeddings. These representations are used for image classification and retrieval, anomaly and out-of-distribution detection, generative models, and time-series or multimodal tasks, with evaluations ranging from ImageNet to UCR and UEA benchmarks. A central open question is how to match latent geometry to the data and downstream task: uniformity can aid retrieval and estimation, while concentrated or class-structured embeddings may better preserve semantic distinctions.

Latest papers

CardsList

26 papers2 papers in the last four weeks, down 50% on the four weeks before. 0.0% of all new papers.

  1. Hyperspherical Simplex Representations for Backward-Compatible Retrieval

    Oct 21, 2026Simone Ricci, Niccolò Biondi, Federico Pernici +1Representation LearningHyperspherical Representation Learning

    Media Integration and Communication Center (MICC) Dipartimento di Ingegneria dell’Informazione University of Florence · Multimedia & Human Understanding (MHUG) Department of Information Engineering and Computer Science University of Trento

  2. Sphere Encoder 2

    Oct 1, 2026Kaiyu Yue, Sean McLeish, Ruchit Rawal +3AutoencodersImage Generation

    University of Maryland · Lawrence Livermore National Laboratory · Cornell University

  3. Understanding Hyperspherical Geometry of ECAPA-TDNN Embedding and Its Impact on Zero-Shot Voice Conversion

    Sep 21, 2026Mathilde Abrassart, Nicolas Obin, Axel RoebelZero-Shot Voice ConversionNeural Representation Geometry

    STMS Lab, IRCAM CNRS, Sorbonne Universit´e Paris, France

  4. Interpretable Multi-Hypersphere Deep Anomaly Detection for Open-set Supervised Anomaly Detection

    Sep 19, 2026Zhiji Yang, Fangyong Wang, Yue Li +2Semi-Supervised Anomaly DetectionAnomaly Detection

    School of Statistics and Mathematics, Yunnan University of Finance and Economics, Kunming, 650221, China · Beijing National Center for Applied Mathematics, Capital Normal University, Beijing, 100048, China

  5. Beyond Gaussian Worlds: Latent Geometry Matters for JEPAs

    Sep 18, 2026Léo Nicollier, Enric Meinhardt-Llopis, Marc Pic +2Joint-Embedding Predictive ArchitectureLatent Space Geometry

    CB, ATT · Université Paris-Saclay, CNRS, ENS Paris-Saclay, Centre Borelli, France · Advanced Track and Trace +5

  6. CNsEMD: An Expert-Annotated Multi-Field-Strength MRI Dataset and a Hyperspherical Manifold Network for Multimodal Cranial Nerve Parcellation

    Sep 7, 2026Lei Xie, Junxiong Huang, Guoqiang Xie +9Cross-Modality Medical Image SegmentationMedical Image Benchmarks

    Advanced Interdisciplinary Science and Technology, Zhejiang University of Technology, Hangzhou 310014, China · College of Information Engineering, Zhejiang University of Technology, Hangzhou 310023, China · Department of Neurosurgery, Nuclear Industry 215 Hospital of Shaanxi Province, Xianyang 712000, China +3

  7. Absorbing Gradient Conflicts: Modeling Semantic Variance via Kent Distributions for Cross-Modal Hashing

    Aug 25, 2026Hengjie Zhu, Dayan Wu, Zihao Zhang +5Cross-Modal RetrievalGradient Interference

    Institute of Information Engineering, Chinese Academy of Sciences · School of Cyber Security, University of Chinese Academy of Sciences · School of Cyber Science and Engineering, Southeast University

  8. Training nGPT

    Aug 2, 2026Ilya Loshchilov, Boris GinsburgHyperspherical Representation LearningEfficient Language Model Training

    NVIDIA

  9. dRAE: Representation Autoencoder with Hyper-Spherical Codes

    Jul 24, 2026Tianren Ma, Lin Long, Chuyan Chen +4AutoencodersHyperspherical Representation Learning

    University of Chinese Academy of Sciences · Ant Group · Zhejiang University +1

  10. The Hyperspherical Geometry of CLIP Latent Space: A Semantic Mixture Model

    Jul 15, 2026Zijie Yu, Gaowen Liu, Ramana Rao Kompella +2Latent Variable ModelsHyperspherical Representation Learning

    Tsinghua University · Cisco Research · University of Illinois Chicago

  11. Expanding SPHERE-JEPA: A Family of Statistical Regularizers for the Hypersphere

    Jun 16, 2026Léo Nicollier, Enric Meinhardt-Llopis, Max Dunitz +3Maximum Mean DiscrepancyHyperspherical Representation Learning

    CB, ATT · Université Paris-Saclay, CNRS, ENS Paris-Saclay, Centre Borelli, France · CB +4

  12. Learning Hyperspherical Time-Frequency Representations for Time-Series Out-of-Distribution Detection

    May 29, 2026Willian T. Lunardi, Samridha Shrestha, Martin AndreoniHyperspherical Representation LearningDistribution Shift Detection

    Technology Innovation Institute, Abu Dhabi, United Arab Emirates · Khalifa University, Abu Dhabi, United Arab Emirates

  13. SPHERE-JEPA: Spherical Prediction with Homogeneous Embeddings

    May 26, 2026L{é}o Nicollier, Max Dunitz, Marc Pic +3Self-Supervised LearningJoint-Embedding Predictive Architecture

    CB,ATT · ATT · CB,IFUMI +1

  14. Neural Collapse by Design: Learning Class Prototypes on the Hypersphere

    May 19, 2026Panagiotis Koromilas, Theodoros Giannakopoulos, Mihalis A. Nicolaou +1Prototype-Based ClassificationHyperspherical Representation Learning

    University of Athens · NSCR Demokritos · University of Cyprus +2

  15. MARGIN: Margin-Aware Regularized Geometry for Imbalanced Vulnerability Detection

    May 11, 2026Yuteng Zhang, Huifang Ma, Jiahui Wei +2Class-Imbalanced LearningSoftware Vulnerability Detection

    College of Artificial Intelligence and Computer Science, Northwest Normal University, Lanzhou 730070, China

  16. Polaris: Coupled Orbital Polar Embeddings for Hierarchical Concept Learning

    Apr 30, 2026Sahil Mishra, Srinitish Srinivasan, Sourish Dasgupta +1Hierarchical ClassificationHierarchical Representation Learning

    Indian Institute of Technology Delhi, New Delhi, India · KDM Lab, Dhirubhai Ambani University, Gandhinagar, Gujarat, India · Indian Institute of Technology Delhi, Abu Dhabi, UAE.

  17. Beyond Gaussian Bottlenecks: Topologically Aligned Encoding of Vision-Transformer Feature Spaces

    Apr 30, 2026Andrew Bond, Ilkin Umut Melanlioglu, Erkut Erdem +1Variational AutoencodersHyperspherical Representation Learning

    Department of Computer Engineering, Kocı University, Istanbul, Turkey · KUIS AI Research Center, Istanbul, Turkey · Department of Electrical and Electronics Engineering, Kocı University, Istanbul, Turkey +1

  18. Hyperspherical Forward-Forward with Prototypical Representations

    Apr 30, 2026Shalini Sarode, Brian Moser, Joachim Folz +4Prototype-Based LearningEfficient Neural Network Inference

    German Research Center for Artificial Intelligence (DFKI) · RPTU Kaiserslautern-Landau

  19. Frequency-Aware Semantic Fusion with Gated Injection for AI-generated Image Detection

    Apr 30, 2026Shuchang Zhou, Shangkun Wu, Jiwei Wei +4Hyperspherical Representation LearningAI-Generated Image Detection

    University of Electronic Science and Technology of China · Harbin Institute of Technology

  20. Self-Supervised Representation Learning via Hyperspherical Density Shaping

    Apr 27, 2026Esteban Rodríguez-Betancourt, Edgar Casasola-MurilloMulti-View LearningHyperspherical Representation Learning

    Posgrado en Computación e Informática · Universidad de Costa Rica · Escuela de Ciencias de la Computación

  21. Neural Precoding in Complex Projective Spaces

    Mar 8, 2026Zaid Abdullah, Merouane Debbah, Symeon Chatzinotas +1Wireless CommunicationsGeometric Representation Learning

    Interdisciplinary Centre for Security, Reliability, and Trust (SnT) at the University of Luxembourg, Luxembourg. · 6G Research Center, Khalifa University of Science and Technology, Abu Dhabi, UAE.

  22. Interpretable Maximum Margin Deep Anomaly Detection

    Mar 7, 2026Zhiji Yang, Mei Huang, Xinyu Li +3Semi-Supervised Anomaly DetectionAnomaly Detection

    the School of Statistics and Mathematics, Yunnan University of Finance and Economics, Kunming 650221, China · the Academy for Multidisciplinary Studies, Capital Normal University, Beijing 100048, China

  23. SphUnc: Hyperspherical Uncertainty Decomposition and Causal Identification via Information Geometry

    Mar 1, 2026Rong Fu, Chunlei Meng, Jinshuo Liu +8Uncertainty DecompositionUncertainty Calibration

    Independent Researcher

  24. In Defense of Cosine Similarity: Normalization Eliminates the Gauge Freedom

    Feb 23, 2026Taha BouhsineHyperspherical Representation Learning

  25. Image Generation with a Sphere Encoder

    Feb 16, 2026Kaiyu Yue, Menglin Jia, Ji Hou +1Image GenerationOne-Step Generative Modeling

    Meta · University of Maryland.

  26. Spherical Cauchy Variational Autoencoders: Heavy Angular Tails and Exact KL Evaluation

    Jun 26, 2025Lukas Sablica, Kurt HornikVariational AutoencodersHyperspherical Representation Learning

    Institute for Statistics and Mathematics Vienna University of Economics and Business Austria