New paper out: Trial-level sequence modeling reveals hidden dynamics of dual-task interference. We decoded, at the trial level, the sequence of cognitive operations people take when multitasking. We found that how cognitive operations interleaved across tasks predicted RT and accuracy!

Very pleased to say that we are (almost) ready for the 1.0.0 of our HMP #Python package to decompose #EEG into single-trial task related events: github.com/GWeindel/hmp There's now a nice documentation: hmp.readthedocs.io/en/latest/we... We just need some beta testers, thus RT appreciated!

Pre-print out arxiv.org/abs/2504.10028 If you decide to click on this URL too quickly, you might just skip a cognitive operation! We combine Hidden Multivariate Pattern analysis (HMP) and deep learning (Mamba/State Space Models) to detect cognitive operations at a trial level from EEG data.

Hello world