Single-peaked tuning curves found in early sensory areas are more optimized for quick decoding than accuracy, while multi-peaked tuning curves (e.g. grid cells) give higher accuracy but only at longer time scales.

The representation of odor transforms from high dimensional, population level encoding of valence and identity in the OT, to a low dimensional representation of valence in the VP.

New computational models provide insights into how the insect brain estimates the speed and direction of movement.

Computational modeling demonstrates that population dynamics of neural calcium activity following stimulus offset are consistent with a network mechanism based on recurrent interactions.

Networks simulations and in vivo imaging suggest a stable backbone of stimulus representation formed by neurons with low population coupling, alongside a flexible substrate of neurons with high population coupling.

The GABA input switching from excitatory to inhibitory during maturation of adult-born dentate granule cells is crucial for their integration into the preexisting circuit.

Building on simple unsupervised matrix factorization techniques, the seqNMF algorithm successfully recovers neural sequences in a wide range of simulated and real datasets.

A 'stimulus-computable' modeling approach resolves apparent inconsistencies between human and monkey lesion data, implicating perirhinal cortex in visual object perception.

A modelling-based analysis reveals a novel type of selectivity for chromatic contrast in a mouse retinal ganglion cell type, and experimental evidence shows that this feature makes this cell type well suited for detecting behaviourally relevant changes in visual context.

Classical theoretical approaches to studying the cerebellar cortex are generalized to a broader set of learning tasks, revealing that the optimal value for the sparsity of granule cell activity is task-dependent.