The FASTIMAGES benchmark
FASTIMAGES is a benchmark for validating methods that detect neural replay non-invasively in humans, built from parallel MEG and fMRI recordings of participants viewing rapid visual sequences with a known ground truth. It evaluates two sequence-detection approaches — TDLM for MEG and SODA for fMRI — showing that each excels in its own modality but struggles when applied across modalities.
Brain decoding in MEG source space using MNE-Python
This blog post will walk you through the steps required for source reconstructed brain decoding using MEG and MNE-Python. It is a copy of https://github.com/skjerns/MEG-source-space-decoding A lot of MEG decoding work is performed on raw sensor data, even though it seems that source based decoding should be superior, based upon the accuracy of fMRI decoding…
“Challenges in Replay Detection by TDLM in Post-Encoding Resting State”
Here’s the summary of our newest preprint, in which we simulated replay and checked how well TDLM can find it. Check the post https://cimh-clinical-psychology.github.io/DeSMRRest-TDLM-Simulation/…
Do LLMs laugh at electric memes?
We know that LLMs have a pretty good “understanding” of the world (or can simulate one, if you don’t like the wording). Previous research has shown that transformer-style models outperform traditional cognitive modelling on many different tasks and even in modelling cognition across species. However, do they understand what humans experience when they look at…