Hyperdimensional computing (HD), also known as vector symbolic architectures (VSA), is an emerging and promising paradigm for cognitive computing. At its core, HD/VSA is characterized by its distinctive approach to compositionally representing information using high-dimensional randomized vectors. The recent surge in research within this field gains momentum from its computational efficiency stemming from low-resolution representations and ability to excel in few-shot learning scenarios. Nonetheless, the current literature is missing a comprehensive comparative analysis of various methods since each of them uses a different benchmark to evaluate its performance. This gap obstructs the monitoring of the field’s state-of-the-art advancements and acts as a significant barrier to its overall progress. To address this gap, this review not only offers a conceptual overview of the latest literature but also introduces a comprehensive comparative study of HD/VSA classification methods. The exploration starts with an overview of the strategies proposed to encode information as high-dimensional vectors. These vectors serve as integral components in the construction of classification models. Furthermore, we evaluate diverse classification methods as proposed in the existing literature. This evaluation encompasses techniques such as retraining and regenerative training to augment the model’s performance. To conclude our study, we present a comprehensive empirical study. This study serves as an in-depth analysis, systematically comparing various HD/VSA classification methods using two benchmarks, the first being a set of seven popular datasets used in HD/VSA and the second consisting of 121 datasets being the subset from the UCI Machine Learning repository. To facilitate future research on classification with HD/VSA, we open-sourced the benchmarking and the implementations of the methods we review. Since the considered data are tabular, encodings based on key-value pairs emerge as optimal choices, boasting superior accuracy while maintaining high efficiency. Secondly, iterative adaptive methods demonstrate remarkable efficacy, potentially complemented by a regenerative strategy, depending on the specific problem. Furthermore, we show how HD/VSA is able to generalize while training with a limited number of training instances. Lastly, we demonstrate the robustness of HD/VSA methods by subjecting the model memory to a large number of bit-flips. The results illustrate that the model’s performance remains reasonably stable until the occurrence of 40% of bit flips, where the model’s performance is drastically degraded. Overall, this study performed a thorough performance evaluation on different methods and, on the one hand, a positive trend was observed in terms of improving classification performance but, on the other hand, these developments could often be surpassed by off-the-shelf methods. This calls for better integration with the broader machine learning literature; the developed benchmarking framework provides practical means for doing so. © The Author(s) 2025.

Reservoir computing advances the intriguing idea that a nonlinear recurrent neural circuit—the reservoir—can encode spatio-temporal input signals to enable efficient ways to perform tasks like classification or regression. However, recently the idea of a monolithic reservoir network that simultaneously buffers input signals and expands them into nonlinear features has been challenged. A representation scheme in which memory buffer and expansion into higher-order polynomial features can be configured separately has been shown to significantly outperform traditional reservoir computing in prediction of multivariate time-series. Here we propose a configurable neuromorphic representation scheme that provides competitive performance on prediction, but with significantly better scaling properties than directly materializing higher-order features as in prior work. Our approach combines the use of randomized representations from traditional reservoir computing with mathematical principles for approximating polynomial kernels via such representations. While the memory buffer can be realized with standard reservoir networks, computing higher-order features requires networks of ‘Sigma-Pi’ neurons, i.e., neurons that enable both summation as well as multiplication of inputs. Finally, we provide an implementation of the memory buffer and Sigma-Pi networks on Loihi 2, an existing neuromorphic hardware platform. 

Time series classification poses significant challenges due to the inherent temporal order of the data points and the existence of sequential dependencies between them. The ROCKET family, featuring methods like MiniROCKET, MultiROCKET, and HYDRA, is currently a leading approach in this domain, leveraging convolution kernels to aggregate temporal features into encodings for linear classifiers. However, these models encode temporal features over short temporal windows and then aggregate them as an unordered set of encodings over the longer temporal window of the entire data sequence. This prevents these models from capturing any longer sequence structure. To address this design drawback, we propose integrating hyperdimensional computing into ROCKET methods to explicitly incorporate temporal order of the short-term features within the entire time series. This approach enhances the discriminative power of encodings generated by MiniROCKET, MultiROCKET, and HYDRA where longer-term structure exists in the data, leading to increased classification performance with minimal computational overhead. More specifically, we introduce a method to represent time series as high-dimensional vectors through multiplicative binding of ROCKET encodings with encodings representing temporal order, applying this approach across various ROCKET methods. Additionally, we explore different high-dimensional vector representations of temporal order, yielding diverse similarity kernels that enhance classification accuracy. Through experiments on synthetic datasets, we highlight the limitations of ROCKET methods in handling temporal dependencies and show how the methods based on hyperdimensional computing overcome these limitations. Furthermore, our extensive experimental evaluation with real-world datasets included in the recent UCR archive, validates the advantages of our approach, consistently achieving classification improvements across all ROCKET methods that integrate hyperdimensional computing. Notably, our best model achieves a relative error rate reduction of over 50% compared to the best ROCKET model on several UCR datasets.

Neuromorphic computing shows promise for advancing computing efficiency and capabilities of AI applications using brain-inspired principles. However, the neuromorphic research field currently lacks standardized benchmarks, making it difficult to accurately measure technological advancements, compare performance with conventional methods, and identify promising future research directions. This article presents NeuroBench, a benchmark framework for neuromorphic algorithms and systems, which is collaboratively designed from an open community of researchers across industry and academia. NeuroBench introduces a common set of tools and systematic methodology for inclusive benchmark measurement, delivering an objective reference framework for quantifying neuromorphic approaches in both hardware-independent and hardware-dependent settings. For latest project updates, visit the project website (neurobench.ai). 

We propose a system for visual scene analysis and recognition based on encoding the sparse, latent feature-representation of an image into a high-dimensional vector that is subsequently factorized to parse scene content. The sparse feature representation is learned from image statistics via convolutional sparse coding, while scene parsing is performed by a resonator network [1]. The integration of sparse coding with the resonator network increases the capacity of distributed representations and reduces collisions in the combinatorial search space during factorization. We find that for this problem the resonator network is capable of fast and accurate vector factorization, and we develop a confidence-based metric that assists in tracking the convergence of the resonator network. 

We introduce residue hyperdimensional computing, a computing framework that unifies residue number systems with an algebra defined over random, high-dimensional vectors. We show how residue numbers can be represented as high-dimensional vectors in a manner that allows algebraic operations to be performed with component-wise, parallelizable operations on the vector elements. The resulting framework, when combined with an efficient method for factorizing high-dimensional vectors, can represent and operate on numerical values over a large dynamic range using resources that scale only logarithmically with the range, a vast improvement over previous methods. It also exhibits impressive robustness to noise. We demonstrate the potential for this framework to solve computationally difficult problems in visual perception and combinatorial optimization, showing improvement over baseline methods. More broadly, the framework provides a possible account for the computational operations of grid cells in the brain, and it suggests new machine learning architectures for representing and manipulating numerical data. 

The “My Seizure Gauge” competition explored the challenge of forecasting epileptic seizures using non-invasive wearable devices without an electroencephalogram. The organizers and the winning team reflect on their experiences. 

Expand & Sparsify is a principle that is observed in anatomically similar neural circuits found in the mushroom body (insects) and the cerebellum (mammals). Sensory data are projected randomly to much higher-dimensionality (expand part) where only few the most strongly excited neurons are activated (sparsify part). This principle has been leveraged to design a FlyHash algorithm that forms similarity-preserving sparse embeddings, which have been found useful for such tasks as novelty detection, pattern recognition, and similarity search. Despite its simplicity, FlyHash has a number of design choices to be set such as preprocessing of the input data, choice of sparsifying activation function, and formation of the random projection matrix. In this paper, we explore the effect of these choices on the performance of similarity search with FlyHash embeddings. We find that the right combination of design choices can lead to drastic difference in the search performance. 

Multilayer neural networks set the current state of the art for many technical classification problems. But, these networks are still, essentially, black boxes in terms of analyzing them and predicting their performance. Here, we develop a statistical theory for the one-layer perceptron and show that it can predict performances of a surprisingly large variety of neural networks with different architectures. A general theory of classification with perceptrons is developed by generalizing an existing theory for analyzing reservoir computing models and connectionist models for symbolic reasoning known as vector symbolic architectures. Our statistical theory offers three formulas leveraging the signal statistics with increasing detail. The formulas are analytically intractable, but can be evaluated numerically. The description level that captures maximum details requires stochastic sampling methods. Depending on the network model, the simpler formulas already yield high prediction accuracy. The quality of the theory predictions is assessed in three experimental settings, a memorization task for echo state networks (ESNs) from reservoir computing literature, a collection of classification datasets for shallow randomly connected networks, and the ImageNet dataset for deep convolutional neural networks. We find that the second description level of the perceptron theory can predict the performance of types of ESNs, which could not be described previously. Furthermore, the theory can predict deep multilayer neural networks by being applied to their output layer. While other methods for prediction of neural networks performance commonly require to train an estimator model, the proposed theory requires only the first two moments of the distribution of the postsynaptic sums in the output neurons. Moreover, the perceptron theory compares favorably to other methods that do not rely on training an estimator model.

This is Part II of the two-part comprehensive survey devoted to a computing framework most commonly known under the names Hyperdimensional Computing and Vector Symbolic Architectures (HDC/VSA). Both names refer to a family of computational models that use high-dimensional distributed representations and rely on the algebraic properties of their key operations to incorporate the advantages of structured symbolic representations and vector distributed representations. Holographic Reduced Representations [321, 326] is an influential HDC/VSA model that is well known in the machine learning domain and often used to refer to the whole family. However, for the sake of consistency, we use HDC/VSA to refer to the field.Part I of this survey [222] covered foundational aspects of the field, such as the historical context leading to the development of HDC/VSA, key elements of any HDC/VSA model, known HDC/VSA models, and the transformation of input data of various types into high-dimensional vectors suitable for HDC/VSA. This second part surveys existing applications, the role of HDC/VSA in cognitive computing and architectures, as well as directions for future work. Most of the applications lie within the Machine Learning/Artificial Intelligence domain; however, we also cover other applications to provide a complete picture. The survey is written to be useful for both newcomers and practitioners.