Neural Representation Geometry

Ask about this topic

Overview

Neural representation geometry asks how learned features are organized by distances, directions, neighborhoods, manifolds, and topology, and how this structure supports prediction, memory, transfer, or behavior. Researchers compare models and brains using methods such as CKA, Procrustes alignment, and relative representations, and characterize spaces with covariance spectra, intrinsic-dimension estimates, persistent homology, and Riemannian geometry; sparse autoencoders and interventions probe feature structure and causal roles. Work spans vision and language models, contrastive embeddings, world models, and neural recordings, with evaluations including retrieval, out-of-distribution transfer, compositional generalization, and brain alignment. A central tension is that measured geometry can change under function-preserving coordinate transformations, while similarity or geometric patterns alone do not establish shared computation, causal importance, or better performance; robust, task-aware measures remain an open need.

Latest papers

CardsList

259 papers37 papers in the last four weeks, up 118% on the four weeks before. 0.2% of all new papers.

  1. NeurEO: dissecting Earth observation embeddings with computational neuroscience

    Oct 29, 2026Thijs L van der Plas, Jacob JW Bakermans, Ioannis N AthanasiadisGeospatial Foundation ModelsNeural Representation Geometry

    Wageningen University & Research, NL · University College London, UK

  2. Non-Zero Radius: The Missing Component of Neural Networks

    Oct 29, 2026Tiansi Dong, Mateja Jamnik, Pietro LiòLogical ReasoningNeuro-Symbolic Reasoning

    Computer Lab, The University of Cambridge · The Alan Turing Institute

  3. The Geometry of Manifold Learning: Singular foliations via Deep Learning models

    Oct 27, 2026Gabriele Novelli, Eliot Tron, Emanuele Latini +1Neural Representation GeometryManifold Learning

    Dipartimento FIM, Università di Modena, via Campi 213/A, Modena, I-41125, Italy. · SAMM, Université Paris 1 Panthéon Sorbonne, 31 rue Baudricourt, Paris, 75013, France. · Dipartimento di Matematica, Università di Bologna, Piazza Porta San Donato 5, Bologna, I-40126, Italy. +1

  4. Neuronal Identity as an Organizational Basis for Analyzing Neural Population Dynamics

    Oct 27, 2026Seungjae Han, Joshua Yedam You, Soi Kim +5Neural DecodingShared Representation Learning

    School of Electrical Engineering, KAIST, Daejeon, Republic of Korea · Department of Brain and Cognitive Sciences, KAIST, Daejeon, Republic of Korea · Graduate School of Data Science, KAIST, Daejeon, Republic of Korea +3

  5. The Information Geometry of Representation Formation

    Oct 25, 2026Sanjukta KrishnagopalNeural Representation GeometryInformation Geometry

    Department of Computer Science, University of California, Santa Barbara

  6. Rethinking Representation Geometry in Supervised Contrastive Learning under Long-Tailed Data

    Oct 23, 2026Joonhyeok Jeong, Jy-yong Sohn, Kibok LeeContrastive LearningSupervised Contrastive Learning

    Department of Statistics and Data Science, Yonsei University

  7. The Geometry of Meaning in Neural Manifolds

    Oct 22, 2026Valentino MaiorcaNeural Representation GeometrySemantic Correspondence

    Department of Computer Science, Faculty of Information Engineering, Informatics, and Statistics, Sapienza University of Rome

  8. When Do Gaussian Priors Fail? A Latent Geometry Investigation of RL Fine-Tuning

    Oct 20, 2026Wang RuLatent Space GeometryDiffusion Model Fine-Tuning

    Department of Chemical Engineering, Tsinghua University

  9. Geometry Regularization for Vanilla NeRF via Self-Supervised Cross-Ray Consistency

    Oct 20, 2026Zongji WangNeural Radiance Fields3D Reconstruction

  10. Learning the Geometry of Materials Space with Trained Neural Tangent Kernels

    Oct 20, 2026Daniel Kaplan, Gabriel Kotliar, Anirvan M SenguptaNeural Representation GeometryComputational Materials Science

    Center for Materials Theory, Department of Physics and Astronomy, Rutgers University, Piscataway 08854 New Jersey USA · Department of Physics and Astronomy, Rutgers University, Piscataway 08854 New Jersey USA · Department of Physics, University of Central Florida, Orlando, Florida 32816, USA +3

  11. Spherical Fixed-Point Reasoners: Adaptive Loop Halting on the Hypersphere

    Oct 19, 2026Arber Zela, Artavazd Maranjyan, Pol Puigdemont +3Adaptive Reasoning in Language ModelsStructured Reasoning in Language Models

    EPFL · Lamarr Institute · Fraunhofer IAIS +1

  12. Multilingual Brain Decoding Reveals Shared Cortical Semantic Geometry

    Oct 17, 2026Yi Guo, Yong Hu, Michael Kwok-Po Ng +1Cross-Lingual Representation AlignmentNeural Representation Geometry

    Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China. · University of Chinese Academy of Sciences, Beijing, China. · The University of Hong Kong, Hong Kong, China. +1

  13. Representations Have Trajectories: Characterizing Temporal Representation Dynamics in Neural Networks

    Oct 17, 2026Shivaprasad B. Gowda, Sparsh MittalLow-Rank Representation LearningNeural Representation Geometry

    Indian Institute of Information Technology, Nagpur · Indian Institute of Technology, Roorkee

  14. Escaping Parameter Space: Tight Generalization Bounds via Representation Quality

    Oct 16, 2026Niclas Göring, Shuofeng Zhang, Branton DeMoss +1Generalization BoundsNeural Representation Geometry

    Rudolf Peierls Centre for Theoretical Physics, University of Oxford, Oxford, UK · Mathematical Institute, University of Oxford, Oxford, UK

  15. A Topological Signature of Learning under Noisy Labels

    Oct 16, 2026Zhongtian Sun, Fan Mo, Prayag Tiwari +1Noisy-Label LearningNeural Representation Geometry

    University of Kent, United Kingdom · University of Oxford, United Kingdom · Halmstad University, Sweden +1

  16. Learning to Be Guided: Shaping Latent Geometry for Classifier-Free Guidance

    Oct 16, 2026Haoming Liu, Junhao Zhu, Shaoyi Zheng +4Latent Space GeometryDiffusion-Based Image Generation

    New York University · New York University Shanghai · Georgia Institute of Technology

  17. Native Geometry vs. Linear Recoverability: How Alignment Changes Platonic Scaling

    Oct 15, 2026Angus Hoy, Austin Lutterbach, Jasper Chong +2Neural Representation GeometryLatent Space Geometry

    EleutherAI SOAR · UniverseTBD · University of Queensland +4

  18. A Closed Form Depth-Width Law for Neural Collapse in Deep Linear Networks

    Oct 15, 2026Pankhil Gawade, Kirsten Fischer, Guy Wolf +1Neural Representation GeometryNeural Scaling Laws

    Institute of AI for Health, Helmholtz Munich · School of Computation, Information and Technology, Technical University of Munich · Université de Montréal +3

  19. A Phenomenology of Neural Geometry in Vision-Language-Action Models

    Oct 14, 2026Mozes Jacobs, Aiden Swann, Mathilde Papillon +16Neural Representation GeometryVision-Language-Action Models

    Goodfire · Harvard University · Stanford University +1

  20. Order-Consistent Latent Geometry for Reliable Short-Horizon Planning in World Models

    Oct 13, 2026Junlin Wu, Wen Fan, Dandan ZhangLatent Space GeometryLatent-Space Planning

    Imperial College London, London, United Kingdom

  21. Geometric Structure Predicts Per-Example Neighborhood Loss Under PCA Truncation

    Oct 13, 2026Sriram Loganathan, Gokul Anand, Ragavan ArivazhaganText Embedding EvaluationEmbedding Compression

    Department of Computer Science, San Jose State University, San Jose, CA, USA

  22. Deep Speech Representations Track Word Timing along a Low-Dimensional Circular Trajectory

    Oct 12, 2026Burin Naowarat, Hao Tang, Sharon GoldwaterNeural Representation GeometrySelf-Supervised Speech Representation Learning

    The Centre for Speech Technology Research, School of Informatics, University of Edinburgh

  23. Binding Phoneme Identity to Position in Neural Models and the Human Brain

    Oct 11, 2026Nafiseh Momeni, Malo Renaudin, Itzhak Fried +1Neural EncodingNeural Representation Geometry

    Université Paris Cité · École Normale Supérieure – PSL · University of California, Los Angeles

  24. The Ball and the Box: Two Geometries of Computation in Superposition

    Oct 8, 2026Xiaoyu Li, Lequan Lin, Dai Shi +3Feature SuperpositionNeural Representation Geometry

    University of New South Wales · University of Sydney · University of Cambridge

  25. A Geometry-Based Capacity Theory for Finite-Feature Associative Memory

    Oct 6, 2026Jianhai Zhang, Donghao Zhang, Pattarawut Charatpangoon +4Associative MemoryInformation Retrieval

    Department of Biomedical Engineering, Schulich School of Engineering, University of Calgary, and the Ron and Rene Ward Centre for Healthy Brain Aging Research, Hotchkiss Brain Institute, University of Calgary

  26. Beyond Perturbation Magnitude: Direction-Dependent Responses in Multimodal Geometric Representations

    Oct 6, 2026Yongsheng Luo, Wengan He, Yu Li +2Neural Representation GeometryMultimodal Robustness

    Zhuhai College of Science and Technology, Zhuhai, China · Zhuhai UNO Technology Co., Ltd., Zhuhai, China

  27. Loss-Invariant Projections as Passive Probes of Learned Representations

    Oct 5, 2026Akshay Chandrasekhar, Pavlo MelnykNeural Representation GeometryVisual Representation Learning

    Linköping University

  28. Anchor Divergence for Semantic Geometry in Contrastive Learning

    Oct 2, 2026Akash Kannan, Kiho Park, Victor VeitchNeural Representation GeometryContrastive Learning

    University of Chicago

  29. Contrastive Neural Embeddings Reveal Individual Traits Beyond Conversational Role

    Oct 2, 2026Hubert Huang, Michelle McCleod, Brendan Ames +1Neural Representation GeometryContrastive Learning

    Stanford University, Palo Alto, CA 94305, USA · University of Alabama, Tuscaloosa, AL 35401, USA · University of Southampton, Southampton SO17 1BJ, UK

  30. Counterfactual Action Evaluation, Observation Bottlenecks, and Representation Geometry in Joint-Embedding Predictive World Models

    Oct 2, 2026Arjun SubramanianWorld Model EvaluationAction-Conditioned World Models

    Computer Science and Artificial Intelligence Laboratory (CSAIL) Massachusetts Institute of Technology