Vector Symbolic Architectures

Ask about this topic

Overview

Vector Symbolic Architectures (VSAs) study how to encode symbols, roles, and relationships in high-dimensional vectors so structures can be composed and retrieved by similarity. Common approaches include holographic reduced representations (HRR and FHRR), binary XOR-and-shift binding, sparse hyperdimensional codes, and geometric representations such as Clifford or spatial semantic pointers. Researchers apply these methods to associative memory, neuro-symbolic reasoning, learned representations, world models, and robotic localization. A central challenge is preserving reliable retrieval and compositional structure under noise, deep recursion, and limited memory while making operations efficient on real hardware.

Latest papers

CardsList

15 papers0 papers in the last four weeks, against 1 the four weeks before. 0.0% of all new papers.

  1. Learning Holographic Reduced Representations with Clifford Variational Autoencoders

    Sep 23, 2026Mohamed Malek Abid, P. Michael FurlongVariational AutoencodersVisual Representation Learning

    Institute of Neuroinformatics, University of Zurich & ETH Zurich · National Research Council Canada & Centre for Theoretical Neuroscience, Systems Design Engineering, University of Waterloo, Waterloo, Canada

  2. Learning a Vector-Symbolic Model for Socio-Cultural Tasks

    Aug 3, 2026Meera Ray, Swapnika Dulam, Christopher L. DancyComputational Cognitive ModelingVector Symbolic Architectures

    Department of Computer Science and Engineering, The Pennsylvania State University W209 Westgate Building, University Park, PA 16802. USA · Department of Industrial and Manufacturing Engineering, The Pennsylvania State University M310 Leonhard Building, University Park, PA 16802. USA

  3. Gram-Space: Structure-Preserving Codebook Compression for Memory-Efficient Neuro-Symbolic AI

    Aug 2, 2026Weilun Wang, Wantong LiMemory-Efficient InferenceVector Symbolic Architectures

    University of California, Riverside

  4. How to Build Marcus's Algebraic Mind: From Minsky's Emotion-Machine Viewpoint

    Jul 18, 2026Hiroyuki Chuma, Kanji Otsuka, Yoichi SatoComputational Cognitive ModelingVector Symbolic Architectures

    Institute of Innovation Research, Hitotsubashi University (Professor Emeritus) · Meisei University (Professor Emeritus) · Shuhari System

  5. How to Build Marcus's Algebraic Mind: From Thagard's Brain--Mind Viewpoint

    Jul 18, 2026Hiroyuki Chuma, Kanji Otsuka, Yoichi SatoComputational Cognitive ModelingHyperdimensional Computing

    Institute of Innovation Research, Hitotsubashi University (Professor Emeritus) · Meisei University (Professor Emeritus) · Shuhari System

  6. Creating Intelligence: A Computational Foundation for AGI

    Jun 30, 2026Peter OvermannComputational Cognitive ModelingHyperdimensional Computing

  7. Recursive Binding on a Budget: Subspace Carving in Order-p Tensor Memories

    Jun 9, 2026Travis Pence, Daisuke Yamada, Vikas SinghMemory-Efficient InferenceVector Symbolic Architectures

    Department of Computer Science, University of Wisconsin-Madison, United States · Department of Biostatistics and Med. Info., University of Wisconsin-Madison, United States

  8. Disentanglement with Holographic Reduced Representations

    Jun 8, 2026Jhonny J. Velasquez Olivera, Christo K. Thomas, Walid SaadDisentangled Representation LearningUnsupervised Representation Learning

    Virginia Tech · Worcester Polytechnic Institute

  9. Geodesic Flow Matching for Denoising High-Dimensional Structured Representations

    May 29, 2026Karim Habashy, Chris EliasmithFlow MatchingRepresentation Learning

    Department of Systems Design Engineering, University of Waterloo, Waterloo, Canada · Center For Theoretical Neuroscience, University of Waterloo, Waterloo, Canada · Department of Philosophy, University of Waterloo, Waterloo, Canada

  10. How to Build Marcus's Algebraic Mind: Algebro-Deterministic Substrate over Galois Fields

    May 20, 2026Hiroyuki Chuma, Kanji Otsuk, Yoichi SatoComputational Cognitive ModelingHyperdimensional Computing

    Institute of Innovation Research, Hitotsubashi University, Kunitachi, Tokyo 186-8603, Japan (Professor Emeritus) · Meisei University, Hino, Tokyo, Japan (Professor Emeritus) · Shuhari System, Tokyo, Japan

  11. Sutra: Tensor-Op RNNs as a Compilation Target for Vector Symbolic Architectures

    May 20, 2026Emma LeonhartLogical ReasoningDifferentiable Programming

  12. Overcoming the Impedance Mismatch: A Theoretical Roadmap for Fusing Foundation Models and Knowledge Graphs

    Apr 21, 2026Sahil Rajesh DhayalkarVector Symbolic ArchitecturesNeuro-Symbolic Reasoning

    Arizona State University

  13. SRMU: Relevance-Gated Updates for Streaming Hyperdimensional Memories

    Apr 16, 2026Shay Snyder, Andrew Capodieci, David Gorsich +1Hyperdimensional ComputingVector Symbolic Architectures

    George Mason University · George Mason University, USA · Neya Robotics +3

  14. HyperSpace: A Generalized Framework for Spatial Encoding in Hyperdimensional Representations

    Apr 16, 2026Shay Snyder, Andrew Capodieci, David Gorsich +1Hyperdimensional ComputingVector Symbolic Architectures

    George Mason University · George Mason University, Fairfax, Virginia, USA · Neya Robotics +3

  15. Geometric Priors for Generalizable World Models via Vector Symbolic Architecture

    Feb 25, 2026William Youngwoo Chung, Calvin Yeung, Hansen Jin Lillemark +3World Model LearningWorld Models

    University of California, Irvine, United States · University of California, San Diego, United States