GPT2MEG: Quantizing MEG for Autoregressive Generation
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2024
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866915758158643200 |
|---|---|
| author | Csaky, Richard van Es, Mats W. J. Jones, Oiwi Parker Woolrich, Mark |
| author_facet | Csaky, Richard van Es, Mats W. J. Jones, Oiwi Parker Woolrich, Mark |
| contents | Foundation models trained with self-supervised objectives are increasingly applied to brain recordings, but autoregressive generation of realistic multichannel neural time series remains comparatively underexplored, particularly for Magnetoencephalography (MEG). We study (i) modified multichannel WaveNet variants and (ii) a GPT-2-style Transformer, autoregressively trained by next-step prediction on unlabelled MEG. For the Transformer, we propose a simple quantization/tokenization and embedding scheme (channel, subject, and task-condition embeddings) that repurposes a language-model architecture for continuous, high-rate multichannel time series and enables conditional simulation of task-evoked activity. Across forecasting, long-horizon generation, and downstream decoding, GPT2MEG more faithfully reproduces temporal, spectral, and task-evoked statistics of real MEG than WaveNet variants and linear autoregressive baselines, and scales to multiple subjects via subject embeddings. Code available at https://github.com/ricsinaruto/MEG-transfer-decoding. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2404_09256 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | GPT2MEG: Quantizing MEG for Autoregressive Generation Csaky, Richard van Es, Mats W. J. Jones, Oiwi Parker Woolrich, Mark Machine Learning Signal Processing Foundation models trained with self-supervised objectives are increasingly applied to brain recordings, but autoregressive generation of realistic multichannel neural time series remains comparatively underexplored, particularly for Magnetoencephalography (MEG). We study (i) modified multichannel WaveNet variants and (ii) a GPT-2-style Transformer, autoregressively trained by next-step prediction on unlabelled MEG. For the Transformer, we propose a simple quantization/tokenization and embedding scheme (channel, subject, and task-condition embeddings) that repurposes a language-model architecture for continuous, high-rate multichannel time series and enables conditional simulation of task-evoked activity. Across forecasting, long-horizon generation, and downstream decoding, GPT2MEG more faithfully reproduces temporal, spectral, and task-evoked statistics of real MEG than WaveNet variants and linear autoregressive baselines, and scales to multiple subjects via subject embeddings. Code available at https://github.com/ricsinaruto/MEG-transfer-decoding. |
| title | GPT2MEG: Quantizing MEG for Autoregressive Generation |
| topic | Machine Learning Signal Processing |
| url | https://arxiv.org/abs/2404.09256 |