Attention reduces noise correlations through enhancing inhibitory feedback in cortical networks.

Implementing neural changes associated with attention in a deep neural network causes performance changes that mimic those observed in humans and macaques.

Prior expectations, induced by statistical learning, affect sensory processing only for attended stimuli, thereby constraining neurocomputational theories that cast perception as an inferential process integrating prior knowledge and sensory evidence.

Attention-grabbing tasks produce rapid opioid-mediated analgesia via the descending pain modulation network to alter spinal neuronal activity.

The brain processes the same multisensory stimulus differently when the way it sounds, as opposed to the way it looks, is useful for making a decision.

Cerebellar degeneration disrupts the use of interval-based temporal cues, but not rhythmic cues, to proactively modulate perceptual sensitivity, implicating the cerebellum in temporal control of visual attention.

A spatially-tuned normalization model accounts for neuronal responses to attended or unattended stimuli that are presented inside the classical receptive field or the surround, and explains various other observations.

Populations of neurons in the macaque visual cortex are subject to shared fluctuations in gain; these signals exhibit anatomical and functional structure, and their variability is diminished under attention.

The pattern of spatial attention preferences in caudate neurons is altered by superior colliculus inactivation, demonstrating that a superior colliculus to basal ganglia link is important for selective attention.

The different laminar profiles observed across the cortical depth for multisensory and attentional influences indicate partly distinct neural circuitries of information-flow control.