The paper introduces a data‑adaptive test statistic for evaluating conditional independence in the graph structure of stationary Gaussian time series, addressing both short‑memory and long‑memory regimes. It provides a finite‑sample Berry–Esseen type Gaussian approximation and validates the testing procedure via block bootstrap, even in ultra‑high‑dimensional settings. The authors also propose a consistency‑enhancing correction, achieving asymptotic consistency in size and power, and demonstrate the method on fMRI data to explore brain functional connectivity.
arXiv:2609.13609v1 Announce Type: cross Abstract: Network analysis for multivariate time series is popular in many fields, from neuroscience to seismology. The inverse spectral density is a common ch...
arXiv:2604. 16926v2 Announce Type: replace-cross Abstract: Electroencephalography (EEG) foundation models have shown strong potential for learning generalizable representations from large-scale neural data, yet their clinical deployment is hindered by distribution shifts across clinical settings, devices, and populations.
arXiv:2606. 31230v1 Announce Type: new Abstract: We study the task of learning the structure of a $d$-sparse Gaussian graphical model on $n$ variables from a single trajectory of Glauber dynamics.
arXiv:2504. 12594v2 Announce Type: replace Abstract: Conditional independence testing is a critical component of feature screening, invariant statistical models, and causal discovery.
The paper introduces DGOTTA, a framework for temporal memory‑aware online test‑time adaptation on dynamic graphs. DGOTTA comprises three modules: temporal‑aware augmentation to diversify test graphs, memory‑aware model prediction to mitigate catastrophic forgetting, and consistency‑guided online adaptation to enforce temporal alignment and smoothness. Experiments on three real‑world datasets and four DGNN backbones show that DGOTTA improves generalization under diverse distribution shifts and across multiple model architectures.
arXiv:2609.37934v1 Announce Type: new Abstract: Forecasting time-varying functional connectivity from electroencephalography (EEG) requires modeling both history-dependent trends and structured varia...
arXiv:2512. 19510v2 Announce Type: replace Abstract: Conditional independence (CI) is central to causal inference, feature selection, and graphical modeling, yet it is untestable in many settings without additional assumptions.
Forecasting time-varying functional connectivity from electroencephalography (EEG) requires modeling both history-dependent trends and structured variability across channels. Conditional flow matching...
arXiv:2607. 18559v1 Announce Type: cross Abstract: Gaussian graphical model selection is usually studied under independent sampling, but in many applications the data arise as a single trajectory of a dependent stochastic process.
Brain4FMs is a unified benchmark for evaluating brain foundation models (BFMs) on both scalp EEG and intracranial EEG (iEEG). It incorporates 17 representative models and 21 public datasets spanning clinical diagnosis, sleep staging, communication, and affective computing, and offers dataset-aware preprocessing, cross‑subject evaluation, and standardized downstream workflows. The benchmark highlights that no single BFM consistently outperforms others across all tasks, modalities, and adaptation protocols, prompting further exploratory analyses of model‑specific spatial, spectral, and discrete representations.
arXiv:2606. 28719v1 Announce Type: new Abstract: Test-time adaptation (TTA) of vision-language models (VLMs) is essential for their robust deployment in dynamic, real-world environments.
NeurDuo-EEG is a causal EEG foundation model that introduces channel‑resolved persistent memory and multi‑timescale memory management, allowing it to model continuous EEG with a fixed‑size state. Trained on 3,955 hours of EEG from 17 public datasets, it outperforms baselines on most short‑window and long‑sequence tasks, notably improving seizure detection AUC‑PR from 0.285 to 0.471. The Small variant achieves these gains with only 4.7 M parameters and supports efficient streaming inference with constant per‑chunk latency even as history grows to one hour.