Pinging the brain to reveal hidden neural states underpinning flexible human cognition

Human cognition is uniquely flexible. We can attend to different aspects of the world and make arbitrary connections between things we see and do, enabling us to perceive, think, and act adaptively. These foundational processes are thought to depend on frontoparietal brain networks, which prioritise relevant sensory information and organise it into temporary, task-specific structures. A key proposal is that flexibility arises from the rapid and transient linking of stimulus and response representations via versatile domain-general cells in these regions. Intriguingly, in this framework, a person’s knowledge of the current task is encoded in the temporary pattern of connections between neurons, rather than in explicit activity. However, studying these essential knowledge states in working brains is difficult, because the corresponding momentary patterns of synaptic potentiation are typically invisible to neuroimaging. PINGSIGHT will develop and use new methods to reveal these hidden neural states. We will inject non-specific ‘pings’ of energy into the system, through visual, auditory, and non-invasive transcranial magnetic stimulation, in the moments before and after active task performance. Then, we will use multimodal neuroimaging and multivariate decoding analyses to read out how the ping activity propagates in the brain, exposing the underlying patterns of synaptic potentiation. We will use this method to study whether and how these previously hidden brain states support two foundational features of flexible cognition: our ability to (1) attend selectively and (2) respond flexibly to the world around us. As well as revealing latent neural mechanisms underpinning these key core processes, there will be important cross-field impacts: exposing the importance of hidden neural states for cognition, and showcasing how to make them visible for interrogation in humans.