Skip to content

Neural State Prediction: Obstructing Shortcut Learning in EEG Foundation Models

Sep 2026 · 0 citations · 72 references
Computer Science

TL;DR

Neural State Prediction (NSP), a latent-predictive framework that constrains both the prediction target and the available context, is introduced and jointly designing latent targets and their context offers a promising direction for EEG foundation models that learn from distributed signal structure.

Abstract

EEG foundation models increasingly use masked prediction to learn from unlabeled recordings, but optimizing this objective does not ensure transferable neural representations. A central challenge is that stable positional cues and local correlations can make masked regions predictable without integrating distributed neural context. To reduce this reliance on low-information prediction paths, we introduce Neural State Prediction (NSP), a latent-predictive framework that constrains both the prediction target and the available context. NSP uses a Target Encoder updated by an exponential moving average (EMA) to define latent supervision. Identity residualization removes additive effects associated with channel identity and relative time from the targets, while topology-separated context excludes their immediate spatial and temporal neighborhood from the visible input. We pretrain NSP on 2.2 million EEG segments from TUEG and evaluate it across 30 downstream datasets spanning clinical diagnosis, sleep staging, emotion recognition, motor imagery, event-related potentials, cognitive-state decoding, and language retrieval. Under full-parameter multi-task fine-tuning on EEG-FM-Bench, NSP achieves 63.94 macro balanced accuracy across 14 datasets, exceeding the strongest evaluated baseline by 2.35 percentage points. Controlled component ablations assess the contribution of each mechanism, while matched context controls and held-out interventions characterize the role of context geometry, signal content, and positional information. Jointly designing latent targets and their context offers a promising direction for EEG foundation models that learn from distributed signal structure.

View source

Similar papers

#machine learning Preprint Oct 2026

NeuroLens: Learning Latent Embeddings of Neural Semantics from Chronic Recordings

NeuroLens (Latent Embeddings of Neural Semantics), a self-supervised model based on the Joint-Embedding Predictive Architecture (JEPA) framework that learns denoised, semantically informative latents from chronic neural recordings, is introduced as a new paradigm for studying how neural representations change during le...

Hanrui Lyu, Bai-Yuan Chen, Tian Tan et al. · 0 citations
#machine learning Preprint Oct 2026

SPERA: Spherical Prior EEG Foundation Model with Geometry- and Frequency-Aware Latent Prediction

Electroencephalography (EEG) provides a non-invasive measure of ongoing neural activity, but building general-purpose EEG models remains challenging due to the heterogeneity of subjects, devices, and electrode montages. Existing EEG foundation models predominantly rely on reconstruction-based objectives defined on the...

Minsu Kim, Ye-Sung Kim, Hye-seon Jeon et al. · 0 citations
Preprint Aug 2026

Predicting Only from Selected Evidence: A Tempered Product-of-Experts Bottleneck for Auditable EEG Diagnosis

Pretrained EEG backbones improve transfer performance, but downstream diagnosis heads remain hard to audit: predictions are made from unrestricted hidden states, whereas explanations are usually produced only after the decision. We introduce tPoE-EIB, an evidence-information bottleneck head for adapting EEG backbones u...

Yinghao Wang, Shu-Jian Yu, Duc Han Le et al. · 0 citations
Open access Sep 2026

SEDAT: a hybrid tokenizer for large EEG models

This study proposes the squeeze-and-excitation (SE)-data-adaptive Gaussian average filtering (DAGAF) adaptive tokenizer (SEDAT), a hybrid framework integrating SE-based spatial aggregation, DAGAF-based signal decomposition, instantaneous-frequency-guided adaptive segmentation, and Fourier-domain resampling into a singl...

Muhammad Zulkifal Aziz, Yue Zhuo, Bin-Wen Huang et al. · 0 citations
Open access Sep 2026

Predictive coding networks capture human neural representations missing in supervised DNNs

Neuroscientific learning theories propose that the brain acquires knowledge by constructing internal world models. Supervised learning, the dominant approach in deep neural networks (DNN), relies on external category labels, making it difficult to reconcile with biological learning. There is an increasing trend towards...

Dirk Gütlin, Denise Kittelmann, Ryszard Auksztulewicz · 0 citations

Related blog posts

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.