Multi-variate time-series are one of the primary data modalities involved in large classes of problems, where deep learning models represent the state-of-the-art solution. In the healthcare domain electrophysiological data, such as intracranial electroencephalography (iEEG), is used to perform a variety of tasks. However, iEEG models require that the number of channels be fixed, while iEEG setups in clinics are highly personalized and thus vary considerably from one subject to the next. To address this concern, we propose a channel-adaptive (CA) architecture that seamlessly functions on any multi-variate signal with an arbitrary number of channels. Each CA-model can be pre-trained on a large corpus of iEEG recordings from multiple heterogeneous subjects, and then finetuned to each subject using equal or lower amounts of data compared to existing state-of-the-art models, and in only 1/5 of the time. We evaluate our CA-models on a seizure detection task both on a short-term ($\sim$15 hours) and a long-term ($\sim$2600 hours) dataset. In particular, our CA-EEGWaveNet - based on EEGWaveNet - is trained on a single seizure of the tested subject, while the baseline EEGWaveNet is trained on all but one. CA-EEGWaveNet surpasses the baseline in median F1-score (0.78 vs 0.76). Similarly, CA-EEGNet - based on EEGNet - also surpasses its baseline (0.79 vs 0.74). Overall, we show that the CA architecture is a drop-in replacement for existing seizure classification models, bringing better characteristics and performance across the board.

Selective state-space models (SSMs) are an emerging alternative to the Transformer, offering the unique advantage of parallel training and sequential inference. Although these models have shown promising performance on a variety of tasks, their formal expressiveness and length generalization properties remain underexplored. In this work, we provide insight into the workings of selective SSMs by analyzing their expressiveness and length generalization performance on regular language tasks, i.e., finite-state automaton (FSA) emulation. We address certain limitations of modern SSM-based architectures by introducing the Selective Dense State-Space Model (SD-SSM), the first selective SSM that exhibits perfect length generalization on a set of various regular language tasks using a single layer. It utilizes a dictionary of dense transition matrices, a softmax selection mechanism that creates a convex combination of dictionary matrices at each time step, and a readout consisting of layer normalization followed by a linear map. We then proceed to evaluate variants of diagonal selective SSMs by considering their empirical performance on commutative and non-commutative automata. We explain the experimental results with theoretical considerations.

All data modalities are not created equal, even when the signal they measure comes from the same source. In the case of the brain, two of the most important data modalities are the scalp electroencephalogram (EEG), and the intracranial electroencephalogram (iEEG). iEEG benefits from a higher signal-to-noise ratio (SNR), as it measures the electrical activity directly in the brain, while EEG is noisier and has lower spatial and temporal resolutions. Nonetheless, both EEG and iEEG are important sources of data for human neurology, from healthcare to brain–machine interfaces. They are used by human experts, supported by deep learning (DL) models, to accomplish a variety of tasks, such as seizure detection and motor imagery classification. Although the differences between EEG and iEEG are well understood by human experts, the performance of DL models across these two modalities remains under-explored. To help characterize the importance of clean data on the performance of DL models, we propose BrainCodec, a high-fidelity EEG and iEEG neural compressor. We find that training BrainCodec on iEEG and then transferring to EEG yields higher reconstruction quality than training on EEG directly. In addition, we also find that training BrainCodec on both EEG and iEEG improves fidelity when reconstructing EEG. Our work indicates that data sources with higher SNR, such as iEEG, provide better performance across the board also in the medical time-series domain. This finding is consistent with reports coming from natural language processing, where clean data sources appear to have an outsized effect on the performance of the DL model overall. BrainCodec also achieves up to a 64x compression on iEEG and EEG without a notable decrease in quality. BrainCodec markedly surpasses current state-of-the-art compression models both in final compression ratio and in reconstruction fidelity. We also evaluate the fidelity of the compressed signals objectively on a seizure detection and a motor imagery task performed by standard DL models. Here, we find that BrainCodec achieves a reconstruction fidelity high enough to ensure no performance degradation on the downstream tasks. Finally, we collect the subjective assessment of an expert neurologist, that confirms the high reconstruction quality of BrainCodec in a realistic scenario. The code is available at https://github.com/IBM/eeg-ieeg-brain-compressor.

This work presents a first evaluation of two state-of-the-art Large Reasoning Models (LRMs), OpenAI’s o3-mini and DeepSeek R1, on analogical reasoning, focusing on well-established nonverbal human IQ tests based on Raven’s progressive matrices. We benchmark with the I-RAVEN dataset and its extension, I-RAVEN-X, which tests the ability to generalize to longer reasoning rules and ranges of the attribute values. To assess the influence of visual uncertainties on these symbolic analogical reasoning tests, we extend the I-RAVEN-X dataset, which otherwise assumes an oracle perception. We adopt a two-fold strategy to simulate this imperfect visual perception: 1) we introduce confounding attributes which, being sampled at random, do not contribute to the prediction of the correct answer of the puzzles, and 2) smooth the distributions of the input attributes’ values. We observe a sharp decline in OpenAI’s o3-mini task accuracy, dropping from 86.6% on the original I-RAVEN to just 17.0%—approaching random chance—on the more challenging I-RAVEN-X, which increases input length and range and emulates perceptual uncertainty. This drop occurred despite spending 3.4x more reasoning tokens. A similar trend is also observed for DeepSeek R1: from 80.6% to 23.2%. On the other hand, a neuro-symbolic probabilistic abductive model, ARLC, that achieves state-of-the-art performances on I-RAVEN, can robustly reason under all these out-of-distribution tests, maintaining strong accuracy with only a modest accuracy reduction from 98.6% to 88.0%. Our code is available at https://github.com/IBM/raven-large-language-models.

Neural networks have shown unprecedented capabilities, rivaling human performance in many tasks. However, current neural architectures are not capable of symbolic manipulation, which is thought to be a hallmark of human intelligence. Vector-symbolic architectures (VSAs) promise to bring this ability through simple vector manipulation, highly amenable to current and emerging hardware and software stacks built for their neural counterparts. Integrating the two models into the paradigm of neuro-vector-symbolic architectures may achieve even more human-like performance. However, despite ongoing efforts, there are no clear guidelines on the deployment of VSA in deep learning-based training situations. In this work, we aim to begin providing such guidelines by offering four practical lessons we have observed through the analysis of many VSA models and implementations. We provide thorough benchmarks and results that corroborate such lessons. First, we observe that Multiply-add-permute (MAP) and Hadamard linear binding (HLB) are up to 3-4$\times$ faster than holographic reduced representations (HRR), even when the latter is equipped with optimized FFT-based convolutions. Second, we propose further speed improvements by replacing similarity search with a linear readout, with no effect on retrieval. Third, we analyze the retrieval performance of MAP, HRR and HLB in a noise-free and noisy scenario to simulate processing by a neural network, and show that they are equivalent. Finally, we implement a hierarchical multi-level composition scheme, with notable benefits to the flexibility of integration of VSAs inside existing neural architectures. Overall, we show that these four lessons lead to faster and more effective deployment of VSA.

Compositional generalization—a key open challenge in modern machine learning—requires models to predict unknown combinations of known concepts. However, assessing compositional generalization remains a fundamental challenge due to the lack of standardized evaluation protocols and the limitations of current benchmarks, which often favor efficiency over rigor. At the same time, general-purpose vision architectures lack the necessary inductive biases, and existing approaches to endow them compromise scalability. As a remedy, this paper introduces: 1) a rigorous evaluation framework that unifies and extends previous approaches while reducing computational requirements from combinatorial to constant; 2) an extensive and modern evaluation on the status of compositional generalization in supervised vision backbones, training more than 5000 models; 3) Attribute Invariant Networks, a class of models establishing a new Pareto frontier in compositional generalization, achieving a 23.43% accuracy improvement over baselines while reducing parameter overhead from 600% to 16% compared to fully disentangled counterparts.

Modern state-space models (SSMs) often utilize structured transition matrices which enable efficient computation but pose restrictions on the model’s expressivity, as measured in terms of the ability to emulate finite-state automata (FSA). While unstructured transition matrices are optimal in terms of expressivity, they come at a prohibitively high compute and memory cost, even for moderate state sizes. We propose a structured sparse parametrization of transition matrices in SSMs that enables FSA state tracking with provably optimal state size and depth, while keeping the computational cost of the recurrence comparable to that of diagonal SSMs. Our method, \emph{PD-SSM}, parametrizes the transition matrix as the product of a column one-hot matrix ($P$) and a complex-valued diagonal matrix ($D$). As a result, the computational cost of parallel scans scales linearly with the state size. Theoretically, the model is BIBO-stable and can emulate any $N$-state FSA with one layer of dimension $N$ and a linear readout of size $N ×N$, significantly improving on all current structured SSM guarantees. Experimentally, the model significantly outperforms a wide collection of modern SSM variants on various FSA state tracking tasks. On multivariate time-series classification, it outperforms neural controlled differential equations, a paradigm explicitly built for time-series analysis. Finally, we integrate PD-SSM into a hybrid Transformer-SSM architecture and demonstrate that the model can effectively track the states of a complex FSA in which transitions are encoded into sets of variable-length English sentences. The code is available at https://github.com/IBM/expressive-sparse-state-space-model.

Classification based on Zero-shot Learning (ZSL) is the ability of a model to classify inputs into novel classes on which the model has not previously seen any training examples. Providing a set of attributes associated with the new class as an auxiliary descriptor is one of the favored approaches to solving this challenging task. In this work, inspired by Hyperdimensional Computing (HDC), we propose the use of stationary distributed binary codebooks in an attribute encoder to compactly represent a computationally simple end-to-end trainable model, which we name Hyperdimensional Computing Zero-shot Classifier (HDC-ZSC). It additionally consists of a trainable image encoder, and a similarity kernel. HDC-ZSC achieves Pareto optimal results with a 63.8 % top-1 classification accuracy on the CUB-200 dataset by having only 26.6 million trainable parameters. Compared to two other state-of-the-art non-generative approaches, HDC-ZSC achieves 4.3% and 9.9% better accuracy, while they require more than 1.85× and 1.72× parameters compared to HDC-ZSC, respectively.

Few-Shot Class-Incremental Learning (FSCIL) enables machine learning systems to expand their inference capabilities to new classes using only a few labeled examples, without forgetting the previously learned classes. Classical backpropagation-based learning and its variants are often unsuitable for battery-powered, memory-constrained systems at the extreme edge. In this work, we introduce Online Few-Shot Class-Incremental Learning (O-FSCIL), based on a lightweight model consisting of a pre-trained and metalearned feature extractor and an expandable explicit memory storing the class prototypes. The architecture is pretrained with a novel feature orthogonality regularization and metalearned with a multi-margin loss. For learning a new class, our approach extends the explicit memory with novel class prototypes, while the remaining architecture is kept frozen. This allows learning previously unseen classes based on only a few examples with one single pass (hence online). O-FSCIL obtains an average accuracy of 68.62% on the FSCIL CIFAR100 benchmark, achieving state-of-the-art results. Tailored for ultra-low-power platforms, we implement O-FSCIL on the 60mW GAP9 microcontroller, demonstrating online learning capabilities within just 12 mJ per new class.

We analyze the capabilities of Transformer language models in learning compositional discrete tasks. To this end, we evaluate training LLaMA models and prompting GPT-4 and Gemini on four tasks demanding to learn a composition of several discrete sub-tasks. In particular, we measure how well these models can reuse primitives observable in the sub-tasks to learn the composition task. Our results indicate that compositional learning in state-of-the-art Transformer language models is highly sample inefficient: LLaMA requires more data samples than relearning all sub-tasks from scratch to learn the compositional task; in-context prompting with few samples is unreliable and fails at executing the sub-tasks or correcting the errors in multi-round code generation. Further, by leveraging complexity theory, we support these findings with a theoretical analysis focused on the sample inefficiency of gradient descent in memorizing feedforward models. We open source our code at https://github.com/IBM/limitations-lm-algorithmic-compositional-learning.

With the advent of deep learning, progressively larger neural networks have been designed to solve complex tasks. We take advantage of these capacity-rich models to lower the cost of inference by exploiting computation in superposition. To reduce the computational burden per input, we propose Multiple-Input-Multiple-Output Neural Networks (MIMONets) capable of handling many inputs at once. MIMONets augment various deep neural network architectures with variable binding mechanisms to represent an arbitrary number of inputs in a compositional data structure via fixed-width distributed representations. Accordingly, MIMONets adapt nonlinear neural transformations to process the data structure holistically, leading to a speedup nearly proportional to the number of superposed input items in the data structure. After processing in superposition, an unbinding mechanism recovers each transformed input of interest. MIMONets also provide a dynamic trade-off between accuracy and throughput by an instantaneous on-demand switching between a set of accuracy-throughput operating points, yet within a single set of fixed parameters. We apply the concept of MIMONets to both CNN and Transformer architectures resulting in MIMOConv and MIMOFormer, respectively. Empirical evaluations show that MIMOConv achieves $\approx 2$–$4\times$ speedup at an accuracy delta within [+0.68, -3.18]% compared to WideResNet CNNs on CIFAR10 and CIFAR100. Similarly, MIMOFormer can handle $2$–$4$ inputs at once while maintaining a high average accuracy within a [-1.07, -3.43]% delta on the long range arena benchmark. Finally, we provide mathematical bounds on the interference between superposition channels in MIMOFormer. Our code is available at https://github.com/IBM/multiple-input-multiple-output-nets.

Continually learning new classes from fresh data without forgetting previous knowledge of old classes is a very challenging research problem. Moreover, it is imperative that such learning must respect certain memory and computational constraints such as (i) training samples are limited to only a few per class, (ii) the computational cost of learning a novel class remains constant, and (iii) the memory footprint of the model grows at most linearly with the number of classes observed. To meet the above constraints, we propose C-FSCIL, which is architecturally composed of a frozen meta-learned feature extractor, a trainable fixed-size fully connected layer, and a rewritable dynamically growing memory that stores as many vectors as the number of encountered classes. C-FSCIL provides three update modes that offer a trade-off between accuracy and compute-memory cost of learning novel classes. C-FSCIL exploits hyperdimensional embedding that allows to continually express many more classes than the fixed dimensions in the vector space, with minimal interference. The quality of class vector representations is further improved by aligning them quasi-orthogonally to each other by means of novel loss functions. Experiments on the CIFAR100, mini-ImageNet, and Omniglot datasets show that C-FSCIL outperforms the baselines with remarkable accuracy and compression. It also scales up to the largest problem size ever tried in this few-shot setting by learning 423 novel classes on top of 1200 base classes with less than 1.6% accuracy drop. Our code is available at https://github.com/IBM/constrained-FSCIL.

With Motor-Imagery (MI) Brain-Machine Interfaces (BMIs) we may control machines by merely thinking of performing a motor action. Practical use cases require a wearable solution where the classification of the brain signals is done locally near the sensor using machine learning models embedded on energy-efficient microcontroller units (MCUs), for assured privacy, user comfort, and long-term usage. In this work, we provide practical insights on the accuracy-cost tradeoff for embedded BMI solutions. Our proposed Multispectral Riemannian Classifier reaches 75.1% accuracy on 4-class MI task. We further scale down the model by quantizing it to mixed-precision representations with a minimal accuracy loss of 1%, which is still 3.2% more accurate than the state-of-the- art embedded convolutional neural network. We implement the model on a low-power MCU with parallel processing units taking only 33.39 ms and consuming 1.304 mJ per classification.

Accurate multiclass classification of electroencephalography (EEG) signals is still a challenging task towards the development of reliable motor imagery brain-computer interfaces (MI-BCIs). Deep learning algorithms have been recently used in this area to deliver a compact and accurate model. Reaching high-level of accuracy requires to store subjects-specific trained models that cannot be achieved with an otherwise compact model trained globally across all subjects. In this paper, we propose a new methodology that closes the gap between these two extreme modeling approaches: we reduce the overall storage requirements by superimposing many subject-specific models into one single model such that it can be reliably decomposed, after retraining, to its constituent models while providing a trade-off between compression ratio and accuracy. Our method makes the use of unexploited capacity of trained models by orthogonalizing parameters in a hyperdimensional space, followed by iterative retraining to compensate noisy decomposition. This method can be applied to various layers of deep inference models. Experimental results on the 4-class BCI competition IV-2a dataset show that our method exploits unutilized capacity for compression and surpasses the accuracy of two state-of-the-art networks: (1) it compresses the smallest network, EEGNet [1], by 1.9×, and increases its accuracy by 2.41% (74.73% vs. 72.32%); (2) using a relatively larger Shallow ConvNet [2], our method achieves 2.95 x compression as well as 1.4% higher accuracy (75.05% vs. 73.59%).

We propose to embed features extracted from event-driven dynamic vision sensors to binary sparse representations in hyperdimensional (HD) space for regression. This embedding compresses events generated across 346×260 differential pixels to a sparse 8160-bit vector by applying random activation functions. The sparse representation not only simplifies inference, but also enables online learning with the same memory footprint. Specifically, it allows efficient updates by retaining binary vector components over the course of online learning that cannot be otherwise achieved with dense representations demanding multibit vector components. We demonstrate online learning capability: using estimates and confidences of an initial model trained with only 25% of data, our method continuously updates the model for the remaining 75% of data, resulting in a close match with accuracy obtained with an oracle model on ground truth labels. When mapped on an 8-core accelerator, our method also achieves lower error, latency, and energy compared to other sparse/dense alternatives. Furthermore, it is 9.84× more energy-efficient and 6.25× faster than an optimized 9-layer perceptron with comparable accuracy.

Motor-Imagery Brain--Machine Interfaces (MI-BMIs)promise direct and accessible communication between human brains and machines by analyzing brain activities recorded with Electroencephalography (EEG). Latency, reliability, and privacy constraints make it unsuitable to offload the computation to the cloud. Practical use cases demand a wearable, battery-operated device with low average power consumption for long-term use. Recently, sophisticated algorithms, in particular deep learning models, have emerged for classifying EEG signals. While reaching outstanding accuracy, these models often exceed the limitations of edge devices due to their memory and computational requirements. In this paper, we demonstrate algorithmic and implementation optimizations for EEGNET, a compact Convolutional Neural Network (CNN) suitable for many BMI paradigms. We quantize weights and activations to 8-bit fixed-point with a negligible accuracy loss of 0.4% on 4-class MI, and present an energy-efficient hardware-aware implementation on the Mr.Wolf parallel ultra-low power (PULP) System-on-Chip (SoC) by utilizing its custom RISC-V ISA extensions and 8-core compute cluster. With our proposed optimization steps, we can obtain an overall speedup of 64x and a reduction of up to 85% in memory footprint with respect to a single-core layer-wise baseline implementation. Our implementation takes only 5.82 ms and consumes 0.627 mJ per inference. With 21.0GMAC/s/W, it is 256x more energy-efficient than an EEGNET implementation on an ARM Cortex-M7 (0.082GMAC/s/W).

In recent years, deep learning (DL) has contributed significantly to the improvement of motor-imagery brain-machine interfaces (MI-BMIs) based on electroencephalography (EEG). While achieving high classification accuracy, DL models have also grown in size, requiring a vast amount of memory and computational resources. This poses a major challenge to an embedded BMI solution that guarantees user privacy, reduced latency, and low power consumption by processing the data locally. In this paper, we propose EEG-TCNet, a novel temporal convolutional network (TCN) that achieves outstanding accuracy while requiring few trainable parameters. Its low memory footprint and low computational complexity for inference make it suitable for embedded classification on resource-limited devices at the edge. Experimental results on the BCI Competition IV-2a dataset show that EEG-TCNet achieves 77.35% classification accuracy in 4-class MI. By finding the optimal network hyperparameters per subject, we further improve the accuracy to 83.84%. Finally, we demonstrate the versatility of EEG-TCNet on the Mother of All BCI Benchmarks (MOABB), a large scale test benchmark containing 12 different EEG datasets with MI experiments. The results indicate that EEG-TCNet successfully generalizes beyond one single dataset, outperforming the current state-of-the-art (SoA) on MOABB by a meta-effect of 0.25.