Neural populations may depend on balanced recurrent connectivity to produce an efficient stimulus representation while also maintaining an accurate stimulus encoding despite the variability introduced by adapting neural responses.

Computational modeling revealed that balanced assemblies of excitatory and inhibitory neurons shape representational manifolds in olfactory cortex-like recurrent networks, resulting in joint maps of sensory and semantic information.

Spiking networks compensate the loss of neurons instantaneously, when restoration of excitatory/inhibitory balance becomes equivalent to restoration of functionality.

Critical networks show maximum pre-stimulus phase and amplitude regulation of stimulus-evoked responses.

Neural oscillations are a necessary consequence of efficient coding of sensory signals by a spiking neural network, limited by synaptic delays and noise.

Combining experimental and theoretical approaches reveals the interplay between optogenetic manipulation and recurrent interactions in mouse cortex.

Default mode network and visual cortex are connected via two parallel pathways that differentially respond to the processing of visual scenes and semantic information about objects, reflecting domain-specific organisation.

The interplay of recurrent excitation and short-term plasticity enables nonlinear transient amplification, an ideal mechanism for selective amplification, pattern completion, and pattern separation in recurrent neural networks.

Short-ranged and random connectivity are sufficient to explain complex, long-range activity patterns observed in macaque motor cortex that are, moreover, flexibly adaptable to behavior.