Having a great time at the @flatironinstitute.org CCN software workshop in San Diego! Do check out their fantastic Python packages Pynapple @pynapple.bsky.social (for general neurophys analysis) and NeMoS (for GLM) github.com/flatironinstitute/nemos - they are an absolute pleasure to work with 🤩

Turns out once you get into ring attractors it's quite hard to give them up... Who knew?!

This one! L. Llewellyn-Jones is a top Welsh historian and a big proponent of what he calls the 'Persian version' (😉) of classical history... Basically relying more on Persian sources (like the inscriptions in Persepolis) than the super-biased Hellenic accounts. Highly recommended!

A 🍍 spotted in the wild at @edinburgh-uni.bsky.social!

Great crash course on neuro data processing with @pynapple.bsky.social and @spikeinterface.bsky.social delivered by @matthiashennig6.bsky.social, @wulfdewolf.bsky.social and Chris Halcrow. So nice to see both libraries working so well together

Fun fact: apparently "dulab" can mean “wheel” in Arabic - seems quite fitting for a lab obsessed with ring manifolds...

Thrilled to announce I'll be joining the Division of Neuroscience at @manchester.ac.uk as a Lecturer in Feb '26!

The DuLab will explore how the brain integrates sensory streams into internal maps 🌐 using the rodent head-direction circuit, the 'neural compass', as a starting point for the journey 🧭

Beyond serving as a neural compass, the thalamocortical HD circuit also shows remarkably coherent activity during sleep we also discuss a few 'smoking guns' that it may support memory-related processing during the nonREM sleep stage... (6/8)

But postsubiculum is more than just a relay! We know that it sharpens the HD tuning inherited from AD and plays an important role in anchoring the HD signal to external landmarks. The actual mechanisms are not yet clear but the microcircuits within the postsubiculum may give us some clues (5/8)

Another argument: activity in postsubiculum is largely driven by inputs from the antrodorsal (AD) nucleus in the thalamus - a nucleus that shows many characteristics of a primary relay (e.g. subcortical driver input) despite being surrounded by higher-order nuclei (4/8)

Why? HD is by far the main behavioural correlate in the postsubiculum: around 50-85% of excitatory cells (depending on who you ask) strongly code for it, in comparison to e.g. ~5-10% in retrosplenial cortex. This is evident in its population activity, which as a whole forms a beautiful 1D ring (3/8)

Despite its name and location, postsubiculum is not just another cogwheel in the hippocampal machinery - it's actually the main conduit of the subcortical head-direction signal to other cortical areas. We propose that it can be conceptualised as the 'primary' cortex for this information stream (2/8)