aiqbench analysis
STEAM introduces a hierarchical transfer framework using soft mixture-of-experts (SSMoE) modules to address generalizability issues in EEG-based brain-computer interfaces, effectively balancing spatial and temporal representations. This approach enables efficient downstream adaptation for specific neural engineering paradigms without requiring full retraining, achieving superior decoding accuracy across multiple datasets while maintaining competitive inference costs.
STEAM:ASpatio-TEmporal Alignment Mixture-of-Experts Model with Hierarchical Pre-training for EEG Decoding
arXiv:2608.02070v1 Announce Type: cross Abstract: Brain-computer interfaces (BCIs) have been widely used in motor rehabilitation, disease diagnosis, and other neural engineering scenarios. However, conventional neural signal decoding algorithms often suffer from limited generalizability and high adaptation costs, motivating recent interest in BCI foundation models. Existing approaches still struggle to jointly achieve general transferability, accurate decoding, and efficient downstream adaptation. We present STEAM, a hierarchical transfer framework that reconciles general-purpose representation learning with paradigm-specific specialization in EEG foundation models. The framework is instantiated as a dual-branch spatio-temporal encoder in which a shared soft mixture-of-experts (SSMoE) module aligns the spatial and temporal branches, allowing complementary representations to exchange information through a compact set of soft slots. Across seven downstream datasets and fourteen evaluation settings, STEAM attains the best average rank among the compared methods at a competitive inference cost measured in FLOPs. Building upon the Stage-I general initialization, the hierarchical pre-training strategy further specializes the model to a target paradigm without retraining from scratch, yielding consistent gains in paradigm-specific decoding accuracy.
Abstract:Brain-computer interfaces (BCIs) have been widely used in motor rehabilitation, disease diagnosis, and other neural engineering scenarios. However, conventional neural signal decoding algorithms often suffer from limited generalizability and high adaptation costs, motivating recent interest in BCI foundation models. Existing approaches still struggle to jointly achieve general transferability, accurate decoding, and efficient downstream adaptation. We present STEAM, a hierarchical transfer framework that reconciles general-purpose representation learning with paradigm-specific specialization in EEG foundation models. The framework is instantiated as a dual-branch spatio-temporal encoder in which a shared soft mixture-of-experts (SSMoE) module aligns the spatial and temporal branches, allowing complementary representations to exchange information through a compact set of soft slots. Across seven downstream datasets and fourteen evaluation settings, STEAM attains the best average rank among the compared methods at a competitive inference cost measured in FLOPs. Building upon the Stage-I general initialization, the hierarchical pre-training strategy further specializes the model to a target paradigm without retraining from scratch, yielding consistent gains in paradigm-specific decoding accuracy.
Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
arXiv:2608.02070 [cs.CV]
(or arXiv:2608.02070v1 [cs.CV] for this version)
https://doi.org/10.48550/arXiv.2608.02070 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 264 words
Related coverage
Welcome Muse Spark Open: open-weight frontier multimodal mixture-of-experts model from Meta
Hugging Face · Oct 15, 2026
STELLAR: Spatio-Temporal Environmental Learning with Latent Alignment and Refinement for Long-Tailed Species Distribution Modeling
arXiv · Jun 9, 2026
CausalMoE: A Billion-Scale Multimodal Foundation Model for Granger Causal Discovery with Pattern-Routed Heterogeneous Experts
arXiv · Jun 12, 2026
LLaTSA: Large Language Model-Aligned General-Purpose Transient Stability Analysis
arXiv · Sep 16, 2026