Mohamady El-Gaby

A cognitive map for value-guided choice in the ventromedial prefrontal cortex Value-based decision-making can be conceptualized as navigation within a cognitive map of choice values. During choice, the ventromedial prefrontal cortex (vmPFC) compositionally builds cognitive maps...

1/n

With Cosyne around the corner, I thought I would:

a) make a thread about our “A cognitive map for value-guided choice in the vmPFC” Cell paper (i'm a bit late). shorturl.at/DBA3H

It's Cosyne time!!

Here are some contributions from my lab and some of my close collaborator labs (Thomas Akam, @melgaby.bsky.social, Steve Kennerley, @sameershethmd.bsky.social )

At #cosyne2026? Check out Sumedha Nalluru’s poster from our lab! Showing that reduced cortical inhibition causes false associations. Friday 13th March from 1:15 - 4:15pm.

Looking forward to :)

This is a game changer. Even structures as complex as a torus can be innate.

All of these posters feature nods to a framework we call duplicated structured representations (DSRs) that seems to unify a broad swathe of PFC data and make cool new predictions. Watch this and @kristorpjensen.bsky.social ‘s space for more on this over the coming year! 6/6

@mathiassablemeyer.bsky.social is taking these ideas one step further and showing how a hierarchical structure can be encoded in MEG activity from humans performing nested sequences. This work elegantly rules in/out different models of hierarchical coding! 5/6

I also want to highlight awesome work from friends in the Behrens lab. @skuechenhoff.bsky.social is finding both fMRI and single unit evidence for two different ways of encoding task progress in the human PFC and MTL! 4/6

Mingyu is asking how these goal progress signals come about in the first place. Turns out ventral hippocampus and PFC interactions might be critical for this. A tour de force of simultaneous hippocampus-PFC neuropixels recordings and optogenetic silencing! 3/6

Adam will show how neurons in the mouse PFC, and both medial and lateral Entorhinal cortices compare when it comes to encoding goal and task progress and how this relates to space and time. Featuring 100s of simultaneously recorded LEC neurons! 2/6

First Cosyne for the Mind Mechanisms lab!
We’re broadly interested in world and goal models in the PFC and temporal lobe. Come chat to @adamloydharris.bsky.social and Mingyu Zhu as they present some new findings tomorrow (Thursday) evening. #Cosyne2026 …🧵 1/6

Excited to see this Version Of Record of my work out in @elife.bsky.social!
elifesciences.org/articles/106...
We investigate the mental representation of geometric shapes in adults and children using fMRI and MEG. Each figure has a video of me explaining the figure: go and read it, or read below.

Neuronal spiking in the mammalian forebrain is dominated by a heterogeneous ground state Neuronal firing patterns have significant spatiotemporal variability with no agreed-upon theoretical framework. Using a combined experimental and mode…

With this one in print, I think I finally earned that PhD... 😅
Presented for the first time at the cosyne when the world ended (March 2020). I'll bring over a summary thread from twitter when it was still twitter...

www.sciencedirect.com/science/arti...

A must read for anyone interested in the social brain.

We review studies showing that when brain areas face similar computational demands in social and non-social context, they perform the same computations. We argue that exaptation (repurposing of traits for new functions) played a key role in brain evolution.

Thrilled to finally share this work! 🧠🔊

Using a new reinforcement-free task we show mice (like humans) extract abstract structure from sound (unsupervised) & dCA1 is causally required by building factorised, orthogonal subspaces of abstract rules.

Led by Dammy Onih!
www.biorxiv.org/content/10.6...

Cosyne Viewing Parties

Visas, costs, care responsibilities, and environmental concerns all limit Cosyne attendance. Luckily, the talks are livestreamed; but watching alone is the high road to an aneurism. Hence: viewing parties! Gather regionally to watch Cosyne talks! More info: shorturl.at/3DHZX.

Hybrid neural–cognitive models reveal how memory shapes human reward learning - Nature Human Behaviour Using artificial neural networks applied to human data, Eckstein et al. show that good models of reinforcement learning require memory components that track representations of the past.

Hybrid neural–cognitive models reveal how memory shapes human reward learning
www.nature.com/articles/s41...

New discovery! Value-based decisions reorganize neural state space. Options are first encoded in orthogonal subspaces Then the selected option rotates into a "readout subspace".
Neural subspace reorganization reflects value-based decision making.
www.biorxiv.org/content/10.6...
#neuroscience

*Multi-region computations in the brain*
When two regions are better than one...
doi.org/10.1016/j.ne...
#neuroskyence

It’s PhD defence season. Here is a short guide on how to ace these.

The hippocampus is not a library, it is a simulation engine.

HPC is known for storing maps of the environment but not so known for generating planned trajectories.

This paper proposes that recurrence in CA3 is crucial for planning.

A🧵with my toy model and notes:

#neuroskyence #compneuro #NeuroAI

Do you need your toes to think?

YESS!

Happy to see this one out 😎!

From your humble toes to your noble neurons, you need all your cells to solve your most important problem :

How to stay alive 😎😅

Elena Gutierrez , @sandra-neuro.bsky.social and @mathiassablemeyer.bsky.social

If you attended our SfN symposium on cognitive maps in the PFC then here’s a recap of what was covered. If you didn’t then here’s what you missed! TLDR: converging evidence for structured representations of the world and task in the PFC in humans NHPs and rodents. Led by @sebves.bsky.social and…

🥳!!NEW PREPRINT!!🥳

We show that the tendency to compress complex social information into priors about social structures becomes more pronounced during adolescence.

osf.io/preprints/ps...

I am soooooo excited to share this work, together with @mkwittmann.bsky.social and @yongling.bsky.social.

In Buzzy Papers we set out to compile some of the most exciting neuroscience papers published in the past two years! A panel of 10 neuroscientists selected the papers from a preliminary list our editorial team assembled by considering publications in top neuroscience journals

Postdoctoral Scientist- Kim and Abdus-Saboor Lab Primary Work Address: Dept of Molecular Biology, Princeton, NJ, 08544 Current HHMI Employees, click here to apply via your Workday account. The Kim Lab at Princeton University and Abdus-Saboor Lab at ...

Come do a joint postdoc with me and @tinakim-neuro.bsky.social, splitting your time between Princeton and Columbia. Come design new activity-dependent labeling enzymes and transcriptional reporters optimized for application in the peripheral nervous system.
hhmi.wd1.myworkdayjobs.com/en-US/Extern...

Our new preprint: The primate hippocampus parses experience into event-aligned neural states, creating a temporal scaffold for organizing behavior.

At #SfN25 ? Come chat at my poster — Wednesday AM (PP14; 414.03).

Different learning algorithms achieve shared optimal outcomes in humans, rats, and mice Animals must exploit environmental regularities to make adaptive decisions, yet the learning algorithms that enabels this flexibility remain unclear. A central question across neuroscience, cognitive science, and machine learning, is whether learning relies on generative or discriminative strategies. Generative learners build internal models the sensory world itself, capturing its statistical structure; discriminative learners map stimuli directly onto choices, ignoring input statistics. These strategies rely on fundamentally different internal representations and entail distinct computational trade-offs: generative learning supports flexible generalisation and transfer, whereas discriminative learning is efficient but task-specific. We compared humans, rats, and mice performing the same auditory categorisation task, where category boundaries and rewards were fixed but sensory statistics varied. All species adapted their behaviour near-optimally, consistent with a normative observer constrained by sensory and decision noise. Yet their underlying algorithms diverged: humans predominantly relied on generative representations, mice on discriminative boundary-tracking, and rats spanned both regimes. Crucially, end-point performance concealed these differences, only learning trajectories and trial-to-trial updates revealed the divergence. These results show that similar near-optimal behaviour can mask fundamentally different internal representations, establishing a comparative framework for uncovering the hidden strategies that support statistical learning. ### Competing Interest Statement The authors have declared no competing interest. Wellcome Trust, https://ror.org/029chgv08, 219880/Z/19/Z, 225438/Z/22/Z, 219627/Z/19/Z Gatsby Charitable Foundation, GAT3755 UK Research and Innovation, https://ror.org/001aqnf71, EP/Z000599/1

paper🚨
When we learn a category, do we learn the structure of the world, or just where to draw the line? In a cross-species study, we show that humans, rats & mice adapt optimally to changing sensory statistics, yet rely on fundamentally different learning algorithms.
www.biorxiv.org/content/10.1...