This looks interesting: in familiar environments all active CA1 pyramidal cells are place cells — the "non-place" ones just have unstable fields. Broadly fits my experience; a shift from earlier work holding that many weren’t place cells. https://www.cell.com/iscience/fulltext/S2589-0042(25)00750-3

Really excited to see this out from the lab & CDB colleagues 🧠✨ A new perspective in Nature Reviews Neuroscience brings together an elegant idea: acetylcholine may help the hippocampus learn by prediction, not just store memories. Worth a read 👇 www.nature.com/articles/s41...

Acetylcholine: a candidate substrate for hippocampal predictive learning? A Perspective by William de Cothi, Sarah Shipley & Caswell Barry @shipleysj.bsky.social @caswell.bsky.social #neuroscience #neuroskyence www.nature.com/articles/s41...

Really excited about this work with the awesome Will de Cothi & @shipleysj.bsky.social, out in Nature Reviews Neuroscience. We propose acetylcholine is an analogue of dopamine — tracking not reward prediction errors but state-transition prediction errors. https://doi.org/10.1038/s41583-026-01058-w

Really excited about this work with the awesome Will de Cothi & Sarah Shipley, just out in Nature Reviews Neuroscience. We propose acetylcholine is an analogue of dopamine — tracking not reward prediction errors but state-transition prediction errors. https://doi.org/10.1038/s41583-026-01058-w

Cracking lab walk in the sunshine today. Whole team out together — exactly what summer's for. 🌞

The QS rankings are a crude, reductive instrument that flattens the rich complexity of a university into a single number. But @ucl.ac.uk is at number 8 this year, and I think that's pretty awesome. It's a brilliant place to work, and the science coming out of here right now is genuinely exciting.

Science improves lives.

People like to dunk on MATLAB now, but it’s hard to overstate how large an advance it was at the time and the impact it has had on scientific computing. RIP Cleve Moler www.nytimes.com/2026/06/11/s...

Nice theory paper: prospective coding and path integration emerging as equilibrium solutions of self-organising networks, no pre-wired attractor required. The forward-shifted activity bump lines up with anticipatory firing in superficial entorhinal cortex. arxiv.org/abs/2606.14649

Interesting Transmitter article by Brian DePasquale on what will be the effect of the increasingly widespread use of LLMs to code up research ideas in neuroscience. Includes a somewhat sceptical commentary from me. 😉

Job klaxon 📣 UCL is hiring two NeuroAI / computational neuroscience specialists (Lecturer or Associate Prof) — a big part of the new NeuroAI centre we're growing in Biosciences. Do consider applying if it's relevant to you. JD & details: https://www.jobs.ac.uk/job/DRX021 Happy to take questions.

We're excited to announce that we're recruiting for new academic posts in #AI for #Biodiversity Science, Future #Food Systems & #Health, #NeuroAI, Computational #Neuroscience, and #Cardiovascular/Metabolic Neuroscience. 📅 Apply by 10 July 2026. www.ucl.ac.uk/life-science...

Something I should have read months ago. Ulanovsky's bats fly long tunnels and reveal CA3 as a single-field coder, CA1 as multi-field. Sparse in, dense out — really rather nice. https://www.nature.com/articles/s41586-026-10537-0

A Brief History of Everyone Who Ever Lived: 10 yrs on. @davidolusoga.bsky.social + I will be chewing the fat @unionchapeluk.bsky.social 20/7: history, prehistory + the return of scientific racism. Pic from 2016: I'm 10yrs craggier/less smug looking now. www.intelligencesquared.com/events/adam-...

It has been super fun to work on this, where we learned that: 1) Mouse PFC is needed for model-based override of habits. 2) Separate PFC populations encode a 'policy' and the current 'value'. 3) The policy and value codes alternate theta sweeps towards the goal. ...and much much more!

Be still my beating heart! Super interesting and important. The Mosers show that when rats switch between familiar rooms, each grid module shifts phase independently — and that differential shift across modules drives global remapping in hippocampal place cells. 🔗 doi.org/10.1101/2025.09.24.677985

Know a brilliant postdoc in neuroscience? Encourage them to apply for the SWC Emerging Neuroscientists Seminar Series - a chance to present to the London neuroscience community. Travel and accommodation covered, no CV’s or references needed. www.sainsburywellcome.org/web/content/...

Closing in two days!

Missed this first time — nice fMRI from Wittkuhn & Schuck: people implicitly learn graph structure as a successor representation, with backward replay visible in visual cortex during brief task pauses. Replay strength tracks the SR, not awareness. https://doi.org/10.1073/pnas.2507516122

Flexible goal learning involves coordinated population activity in dCA1 and medial orbitofrontal cortex journals.plos.org/plosbiology/...

Neuropixels + Optogenetics = Neuropixels Opto Combining high-resolution electrophysiology and optogenetics. 960 sites, 28 emitters, 2 colors. Today in @natmethods.nature.com doi.org/10.1038/s415... Thanks to @wellcometrust.bsky.social, @alleninstitute.org, @hhmijanelia.bsky.social & team.

Really nice review of vector coding in the subiculum from Colin Lever — boundary and object vector tuning pulled together in one place. One for the BVC aficionados. https://doi.org/10.1002/hipo.70074

This paper is absolutely wild and amazing. www.science.org/doi/10.1126/...

What do place cells do when the reward keeps moving? Reorganise around the uncertainty. Great task. Also Mellor, Dayan & Dombeck — come on! https://doi.org/10.1101/2025.01.06.631465

Really interesting-looking paper from the Balasubramanian lab — a single recurrent network trained on predictive coding spontaneously produces both grid and place cells. Looking forward to reading it. https://arxiv.org/abs/2605.21356v1

For our next online seminar, we are excited to welcome @noemielteto.bsky.social from DeepMind, who will talk about: "ATLAS: Active Theory Learning for Automated Science" 🗓️Wed 20 May 2026 ⏰2-3pm BST via zoom Mark your date and register for your attending: www.eventbrite.co.uk/e/ucl-neuroa...

Nice preprint from Lee lab — evidence for a link between hippocampal replay and planning has always been thin. First clear evidence replay tracks goals several steps ahead, across subtasks, in rats on a sequential working memory task. https://www.biorxiv.org/content/10.64898/2025.12.02.691753v1

This looks great but I’m getting strong weeping angels vibes.

Had so much fun at Pint of Science last night. Brilliant talk from Holly Gregory. If you missed out last night there’s another night of talks tonight!

Last three tickets for tonight's Pint of Science. Get them while they are hot.

Join our team at the London @pintofscience.uk events. You'll get to hear from @caswell.bsky.social and Holly Gregory about how the body works! Still a few tickets for tomorrow! Book now! pintofscience.co.uk/event/maps-m...

📢 Nominations are now open for the UCL Alumni Awards. Our alumni are making an impact worldwide — advancing ideas, driving change, and supporting the UCL community. The @UCLAlumni Awards recognise these achievements and the people behind them. Submit a nomination by 30 June. https://bit.ly/4dgLSgp

Do you live near London? Are you free on Tuesday night? Do you want to see me or Holly talk about fun neuro things? Well then ... come to Pint Of Science (tickets nearly sold out)

This will be totally brilliant - online talk so available for all! Please come along

This will be so much fun (and a great chance to learn about real science - REALLY REALLY good stuff)

This is cool and potentially very important ...

Wow this looks really cool. Would love to hear more.

New research from @profafia.bsky.social at the UCL School of Pharmacy. Human stem‑cell microcircuit models Alzheimer’s hyperexcitability with patient cells—an early human platform to study mechanisms and test therapies. Read more: www.ucl.ac.uk/life-science... @alzheimersresearchuk.org

Lovely paper on multi-room navigation — RSC has cells whose tuning rotates with each room's geometry, while hippocampal place cells simply remap or translate. Different solutions to encoding repeated structure across complex spaces. https://doi.org/10.1101/2024.08.22.609122

Missed this when it came out last year. Peng et al. find that grid cells in mice running a self-motion task don't hold a single global pattern — instead they reanchor to task-relevant objects, encoding position in multiple local reference frames. https://www.nature.com/articles/s41593-025-02054-6

I'm hosting a #PintOfScience night on the 19th May! Come along for a fun evening with awesome talks from UCL's @caswell.bsky.social and Holly Gregory! @ucllifesciences.bsky.social pintofscience.co.uk/event/maps-m...

Not about the HPC but still cool — preprint comparing representational drift in visual cortex with the same phenomenon in artificial neural networks. The two look strikingly similar. https://doi.org/10.1101/2025.11.05.686897

Our next workshop will be on May 14th - hosted by @flor-iacaruso.bsky.social Come along for some great discussion on building better experiment-theory collaborations in neuroscience! Registration is free, but required

Get a pint with your science!

Awesome opportunity here!

Read it and weep visual cortex!

Great news indeed - at least some common sense has prevailed.

Nicely contrary to how I'd thought this worked: disrupting theta-timescale spiking impairs learning but leaves hippocampal replay intact. Suggests replay doesn't need clean theta sequences during behaviour to form. https://doi.org/10.1101/2025.09.15.675587

What draws researchers into the world of #biodiversity science? 🎥 Hear from our academics about the moments that sparked their passion for understanding and protecting the #natural world. youtu.be/9BT3bL13K4s #EarthDay #EarthDay2026