Search arXiv⌕ Search
Home/Archive/arXiv

arXiv · 2610.00746

MEG-Mamba: A Scalable State-Space Foundation Model for Magnetoencephalography

Chetan Gohil·SungJun Cho·Oiwi Parker Jones·Mark Woolrich
View original source ↗Read source document ↗Export metadata ↓

Abstract

Magnetoencephalography (MEG) is an imaging technique that offers a non-invasive, millisecond-resolution view of human brain activity. The increasing availability of MEG data presents an opportunity to take advantage of a recent advance in artificial intelligence, namely self-supervised foundation models. Existing foundation models for MEG (and electroencephalography) have been built on a transformer architecture. Here, we introduce MEG-Mamba: a generative foundation model for neural activity (source reconstructed, parcellated MEG) built on a Mamba architecture. MEG-Mamba is trained to predict the next token of a discretised MEG signal, conditioned on a brain region and recording session. MEG-Mamba surpasses the generative fidelity of a transformer-based alternative (MEG-GPT) while pre-training in less time (22 vs 400 GPU-hours) and modelling a longer context (4 vs 0.32 s). We evaluate MEG-Mamba by examining the spatio-spectral characteristics of the data it generates and by interpreting its learned embeddings. Furthermore, we demonstrate that lightweight stimulus conditioning (LoRA) can be used to generate realistic task-evoked responses that are not included in pre-training. Our results suggest that Mamba is a promising backbone for scaling neural foundation models.

Explore related subjects

q-bio.NC

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Chetan Gohil, SungJun Cho, Oiwi Parker Jones, Mark Woolrich. 2026-09-30. MEG-Mamba: A Scalable State-Space Foundation Model for Magnetoencephalography. https://arxiv.org/abs/2610.00746

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

One Inference, Four Failure Modes: Formal Models of Why Pain Location Fails

Patient-reported pain location is diagnostically decisive for some presentations and nearly uninformative for others. A companion paper argues this is not one gradient of diagnostic utility but three distinct failures of localization. This paper gives those failures their mathematics and shows they are one object: a single Bayesian generative model failing at different nodes - the likelihood, the model class, and group- or context-dependence in that same likelihood. The count is not in dispute: the first three are failures of the inference that produces a felt location, and the fourth, added here, is a failure of reporting it. Anatomical multiplexing is a non-identifiable inverse problem: a rank-deficient referral matrix sends distinct causes to one report. Delocalized amplification is a change of generative model whose dynamics are a neural-field bifurcation, with spatial extent as the order parameter. Referred and atypical displacement is group- or context-dependence in the presentation likelihood - a probability, not a loss - whose consequence sits one layer downstream, in a decision threshold varying with group prevalence and cost. Repeated observation cannot reduce recoverable information about a fixed inferential target, an exact chain-rule identity, while practical value can fall. The fourth node is the report itself. A spatial Bayesian model factorizes mislocalization into referral blur, an anatomical offset and a precision-weighted cognitive override, and reproduces phantom-limb, mirror-box and central-post-stroke reports as regimes of one equation. It is estimable under conditions the paper states as necessary or sufficient rather than assuming them; Sec. 8 gives the design that meets them. One control is reported against the paper's own interest: the transport statistic first used to measure migration scores a stationary, spreading profile as though it had travelled.

q-bio.NC↗
arXiv2026

A High-Density EEG Dataset for Stimulus-Driven Auditory Attention

Stimulus-driven auditory attention determines which sound gains priority when multiple sources compete without an explicit listening goal, yet most computational studies focus either on acoustic salience or on decoding predefined attended targets. This study investigates instruction-free auditory competition using the Stimulus-driven Auditory Attention (SAAD) paradigm and develops a neurophysiologically informed framework that integrates stimulus-derived sound priority with trial-specific EEG evidence. Behavioral analysis using a Bradley--Terry model showed that sound priority estimated from previous competitions generalized to unseen sound pairings, improving held-out prediction from an AUC of 0.577 to 0.718. EEG analysis further revealed mid-to-late centro-temporal lateralization associated with the reported selection side, with neural information remaining predictive beyond acoustic asymmetry. Guided by these findings, the proposed model first estimates a latent priority for each competing sound and forms relative stimulus evidence from their difference. A multi-scale EEG pathway with complementary signed and power-based readouts then extracts trial-specific neural evidence, which is incorporated through gated decision-level integration. The framework is evaluated using mirror-constrained and pairing-held-out protocols, together with representative acoustic, EEG, multimodal baselines, and systematic ablations. The results support a computational account in which spontaneous auditory selection reflects the interaction between generalizable stimulus priority and trial-specific neural variability.

q-bio.NC↗
arXiv2026

Neural spikes as rare events

We consider the information transmission problem in neurons and its possible implications for learning in neural networks. Our approach is based on recent developments in statistical physics and complexity science. Combining sensory information from various modalities for perceptual decision-making offers several advantages and is essential for the survival of both humans and animals. Not much is known about which brain regions are involved in spatial localization using audiovisual integration. We explore this further by training mice in a task requiring audiovisual integration. We then record from the secondary motor cortex (M2) using high-density electrophysiology. Analyzing this data, we found neurons responsive to multimodal as well as unimodal auditory and visual stimuli. The neurons are generally more responsive to auditory, rather than visual, stimuli. There was low correlation between the auditory and visual responses. Some neurons were sensitive to the task mode, whether active or passive, with more neurons being responsive in the active mode. A relatively large percentage of neurons (10-11%) differed significantly in their response to left and right-sided auditory stimuli, but only in 1 of the 3 mice we recorded from. These findings suggest a role for M2 in multisensory decision making and should enable further research in this field. We then use branching process simulations to model neural activity. This would support temporal coding theory as a model for neural coding.

q-bio.NC↗