A new preprint from the Akrami lab at the Sainsbury Wellcome Centre shows that mice, like humans, can quickly learn patterns in sounds, including frequency and abstract rules, simply from experience, without instruction or reward. Their work demonstrates that the hippocampus performs this function of true unsupervised statistical learning, by dynamically reorganising its activity in separate subspaces, offering new insight into one of the brain’s core computational abilities.

Human cognitive abilities differ from those of other species, but finding clear, observable markers of the gap remains challenging. One domain where this difference becomes visible is our capacity to recognise 2D shapes. While many animals can recognise flat depictions of objects, being able to distinguish plain old squares, rectangles, and triangles seems to be an innate, and uniquely human, ability. Researchers are now investigating the neural mechanisms that make this skill possible.

The longer we stay awake, the stronger the homeostatic pressure to sleep becomes. But what are the precise signals of this need inside the brain, and how does that signal translate into the neural activity that drives rest? Recent studies using fruit flies, undertaken by University of Oxford researchers Dr Raffaele Sarnataro & Dr Peter S. Hasenhuetl, are providing new insights into how sleep is controlled. As two winners of the Emerging Neuroscientist Seminar Series 2025, they recently spoke at SWC about their work. In this article, they introduce their findings on how cellular energy balance and network-level oscillations can regulate sleep.