Abstract Working memory (WM), short term maintenance of information for goal directed behavior, is essential to human cognition. Identifying the neural mechanisms supporting WM is a focal point of neuroscientific research. One prominent theory hypothesizes that WM content is carried in a dynamic fashion, involving an “activity-silent” brain state based on synaptic facilitation. Information carried in such activity-silent brain states could be decodable from content-specific changes in responses to unrelated “impulse stimuli”. A potential method for such impulses is single-pulse transcranial magnetic stimulation (TMS) with its focal, precise nature. Here, we tested the activity-silent model by combining TMS/EEG and multivariate pattern analysis (MVPA) with a non-conceptual auditory WM task that employed parametric ripple sound stimuli and a retro-cue design. Our MVPA employed between-subject cross-validation and robust non- parametric permutation testing. The decoding accuracy of WM content significantly increased after a single pulse TMS was delivered to the posterior superior temporal cortex during WM maintenance. Our results are compatible with the theory that WM maintenance involves brain states which are effectively “activity-silent” relative to other intrinsic processes visible in the EEG signal. Single-pulse TMS combined with MVPA could provide a powerful way to decode information content of “activity-silent” brain states involved in WM.

Decoding auditory working memory content from EEG aftereffects of auditory-cortical TMS