DynamicTRF

The code for Dynamic modeling of EEG responses to natural speech reveals earlier processing of predictable words

This work depends on nnTRF.

Roadmap

🔜 Planned | 🚧 In Progress | 🧪 Testing | ✅ Completed

v0.0.1

✅ Refactor the code while reproducing results in the paper

v1.0.0

🚧 remove dependency on the old mTRFpy
🔜 remove dependency on the old stimrespflow trainer
🔜 easier to use method for creating torch dataset required by the model, with code examples
🔜 switched to light-weight stimrespflow library, with light-weight trainer
🔜 user-friendly function to start dynamic TRF analysis, with just one call

Installation

    pip install git+https://github.com/powerfulbean/DynamicTRF.git

Citing DynamicTRF

Dou, J., Anderson, A. J., White, A. S., Norman-Haignere, S. V., & Lalor, E. C. (2025). Dynamic modeling of EEG responses to natural speech reveals earlier processing of predictable words. PLOS Computational Biology, 21(4), e1013006.

@article{dou2025dynamic,
  title={Dynamic modeling of EEG responses to natural speech reveals earlier processing of predictable words},
  author={Dou, Jin and Anderson, Andrew J and White, Aaron S and Norman-Haignere, Samuel V and Lalor, Edmund C},
  journal={PLOS Computational Biology},
  volume={21},
  number={4},
  pages={e1013006},
  year={2025},
  publisher={Public Library of Science San Francisco, CA USA}
}