用大模型预测脑电数据，实现跨数据集长时间生成。
Scaling Next-Brain-Token Prediction for MEG
- 用改进的向量量化器将多通道脑电转为序列，训练大模型预测下一脑区活动。
- 在超过500小时数据上训练，生成长达数分钟的脑电信号仍保持稳定。
- 适合神经科学与生成模型研究者，尤其关注脑信号建模与跨设备泛化。
我们提出一个大规模自回归模型，用于源空间脑磁图（MEG）的下一个脑区令牌预测，可跨数据集与扫描仪扩展长上下文建模：处理了超过500小时、数千次会话的三大最大MEG数据集。通过修改SEANet风格的向量量化器，将多通道MEG压缩为扁平化的令牌流，并从头训练Qwen2.5-VL骨干网络以预测下一个脑区令牌，递归生成长达数分钟的MEG信号，仅需约1分钟的上下文。为评估长时程生成性能，引入任务匹配测试：(i) 基于生成信号自身漂移与真实滑动窗口时间分辨分布的对比，评估在线流稳定性；(ii) 通过神经生理学基础度量集，比较正确上下文与提示替换控制下的条件特异性。模型在CamCAN和Omega上训练，所有分析在预留的MOUS数据集上进行，验证了跨数据集泛化能力。各项指标显示，生成信号在长滚动中保持相对稳定，且更接近真实后续序列而非替换控制。代码已公开：https://github.com/ricsinaruto/brain-gen。
原文摘要 · Abstract (English)
We present a large autoregressive model for source-space MEG that scales next-token prediction to long context across datasets and scanners: handling a corpus of over 500 hours and thousands of sessions across the three largest MEG datasets. A modified SEANet-style vector-quantizer reduces multichannel MEG into a flattened token stream on which we train a Qwen2.5-VL backbone from scratch to predict the next brain token and to recursively generate minutes of MEG from up to a minute of context. To evaluate long-horizon generation, we introduce task-matched tests: (i) on-manifold stability via generated-only drift compared to the time-resolved distribution of real sliding windows, and (ii) conditional specificity via correct context versus prompt-swap controls using a neurophysiologically grounded metric set. We train on CamCAN and Omega and run all analyses on held-out MOUS, establishing cross-dataset generalization. Across metrics, generations remain relatively stable over long rollouts and are closer to the correct continuation than swapped controls. Code available at: https://github.com/ricsinaruto/brain-gen.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。