Santiago Rompani's Avatar

Santiago Rompani

@santiagorompani

Neuro group leader at EMBL Rome. My lab studies how vision is shaped by cross-modal and state-dependent modulation, among other things.

1,153
Followers
818
Following
58
Posts
27.09.2023
Joined
Posts Following

Latest posts by Santiago Rompani @santiagorompani

We think cortex might function like a JEPA. It looks like prediction errors in layer 2/3 are not computed against input (as is the idea in predictive processing), but against a representation in latent space (i.e. like in a JEPA arxiv.org/abs/2301.08243 or RPL doi.org/10.1101/2025...).

30.01.2026 14:51 πŸ‘ 44 πŸ” 12 πŸ’¬ 2 πŸ“Œ 2
A functional influence based circuit motif that constrains the set of plausible algorithms of cortical function There are several plausible algorithms for cortical function that are specific enough to make testable predictions of the interactions between functionally identified cell types. Many of these algorithms are based on some variant of predictive processing. Here we set out to experimentally distinguish between two such predictive processing variants. A central point of variability between them lies in the proposed vertical communication between layer 2/3 and layer 5, which stems from the diverging assumptions about the computational role of layer 5. One assumes a hierarchically organized architecture and proposes that, within a given node of the network, layer 5 conveys unexplained bottom-up input to prediction error neurons of layer 2/3. The other proposes a non-hierarchical architecture in which internal representation neurons of layer 5 provide predictions for the local prediction error neurons of layer 2/3. We show that the functional influence of layer 2/3 cell types on layer 5 is incompatible with the hierarchical variant, while the functional influence of layer 5 cell types on prediction error neurons of layer 2/3 is incompatible with the non-hierarchical variant. Given these data, we can constrain the space of plausible algorithms of cortical function. We propose a model for cortical function based on a combination of a joint embedding predictive architecture (JEPA) and predictive processing that makes experimentally testable predictions. ### Competing Interest Statement The authors have declared no competing interest. Swiss National Science Foundation, https://ror.org/00yjd3n13 Novartis Foundation, https://ror.org/04f9t1x17 European Research Council, https://ror.org/0472cxd90, 865617

Our work with @georgkeller.bsky.social on testing predictive processing (PP) models in cortex is out on biorvix now! www.biorxiv.org/content/10.6... A short thread on our findings and thoughts on where we should move on from PP below.

30.01.2026 14:37 πŸ‘ 47 πŸ” 16 πŸ’¬ 2 πŸ“Œ 1
Preview
Agouti integrates environmental cues to regulate paternal behaviour - Nature Expression of agouti signalling protein in neurons in the medial preoptic area is increased by group housing and negatively associated with care, and overexpression of Agouti reduces care and enhances...

Why are some males caring toward infants while others are neglectful or abusive? I'm so pleased to share work that my colleagues and I @princeton.edu have just published @nature.com (an explanatory thread to follow!) (1/8)

18.02.2026 17:05 πŸ‘ 140 πŸ” 59 πŸ’¬ 9 πŸ“Œ 7

And save >80% on the cost by pooling 6 or more plasmids into a single sample using our free SAVEMONEY algorithm!

elifesciences.org/articles/88794

colab.research.google.com/github/Masaa...

15.02.2026 03:04 πŸ‘ 101 πŸ” 34 πŸ’¬ 8 πŸ“Œ 2

Ooops forgot to tag my job advert for interns to join my lab with the relevant hashtags:

#NeuroJobs #NeuroSkyence

14.02.2026 15:29 πŸ‘ 2 πŸ” 1 πŸ’¬ 0 πŸ“Œ 0
Video thumbnail

You know what? For €25, this is not a bad way to add a digital read-out to a stereotaxic frame.

I'm measuring about a 20 micron discrepancy per millimeter between the DRO and the stereotax; not terrible.

#neuroSkyence 🧠

13.02.2026 15:15 πŸ‘ 20 πŸ” 3 πŸ’¬ 2 πŸ“Œ 1

My lab is looking for an intern interested in the molecular basis of cognitive decline. Work includes deep phenotyping of animal behavior and downstream analysis, including DLC tracking, motif mining (e.g. with Keypoint-MoSeq), and -omics. On-site at EMBL Rome, paid, 1-year commitment. DMs open.

13.02.2026 13:10 πŸ‘ 6 πŸ” 2 πŸ’¬ 1 πŸ“Œ 0

Baseline activity of V1 interneurons connects pupil-linked arousal to engaged behavioral state https://www.biorxiv.org/content/10.64898/2026.02.05.704022v1

08.02.2026 00:15 πŸ‘ 3 πŸ” 2 πŸ’¬ 0 πŸ“Œ 0

Here is the updated link (sorry for the misdirect):

www.nature.com/articles/s41...

24.01.2026 02:32 πŸ‘ 16 πŸ” 8 πŸ’¬ 1 πŸ“Œ 0
Post image Post image

We could all use some extra stress #resilience rn. But how does it work in the brain? On the surface, resilience looks like ignoring stress. Does the brain fail to respond? Or are active adaptations required? Or are resilience and susceptibility divergent paths? doi.org/10.1016/j.celrep.2025.116867

24.01.2026 13:14 πŸ‘ 60 πŸ” 19 πŸ’¬ 6 πŸ“Œ 0
Preview
1000 Hurts Psychophysics is a human-facing science with interventions arguably more robust than medicine.

Psychophysics is a human-facing science with interventions arguably more robust than medicine.

15.01.2026 15:31 πŸ‘ 66 πŸ” 20 πŸ’¬ 2 πŸ“Œ 6
Preview
Glutamate indicators with increased sensitivity and tailored deactivation rates - Nature Methods iGluSnFR4f and iGluSnFR4s are the latest generation of genetically encoded glutamate sensors. They are advantageous for detecting rapid dynamics and large population activity, respectively, as demonst...

iGluSnFR4 is now available @natmethods.nature.com

www.nature.com/articles/s41...

Glutamate indicators with increased sensitivity and tailored deactivation rates πŸ§ͺ

23.12.2025 13:21 πŸ‘ 39 πŸ” 17 πŸ’¬ 0 πŸ“Œ 0
Post image

This afternoon at #nvp2025, an entire session on cognitive #pupillometry πŸ‘οΈ with talks by @anavili.bsky.social , @cstrauch.bsky.social and others. See you at πŸ‘‰ 12:00! #psychology

18.12.2025 08:36 πŸ‘ 9 πŸ” 4 πŸ’¬ 0 πŸ“Œ 2
Post image

Planning a one-afternoon visual psychophysics practical for undergraduate students. Does anyone know a good platform to collect quantitative data on visual acuity, the blindspot, colour mixture, depth perception, edge detection, lateral inhibition, motion adaptation, visual illusions etc?

18.12.2025 08:15 πŸ‘ 6 πŸ” 1 πŸ’¬ 1 πŸ“Œ 0
Preview
Locus coeruleus norepinephrine neurons facilitate orbitofrontal cortex remapping and behavioral flexibility Ogg et al. use in vivo imaging techniques to record activity in the locus coeruleus (LC) and the orbitofrontal cortex (OFC) during a reversal learning task in freely moving rodents. They show that man...

Thrilled this is out, led by former PD Cameron Ogg (now with her own lab at @rhodescollege.bsky.social!). It was a driven by a desire to see, in real-time, how LC activity/NE release influences downstream targets in behaving animals. SO hard to do, but Cameron did it! www.cell.com/cell-reports...

15.12.2025 14:47 πŸ‘ 63 πŸ” 31 πŸ’¬ 6 πŸ“Œ 0

Absence of Local Retinotopy in the Mouse Optic Tract https://www.biorxiv.org/content/10.1101/2025.11.23.690078v1

26.11.2025 17:15 πŸ‘ 2 πŸ” 1 πŸ’¬ 0 πŸ“Œ 0
Preview
Without monkeys, neuroscience has no future Research in primate brains has been essential for the development of BCIs, ANNs. New funding and policy changes put future such advances at risk.

Research in primate brains has been essential for the development of brain-computer interfaces and artificial neural networks. New funding and policy changes put future such advances at risk, write Cory Miller, @movshon.bsky.social and Doris Tsao.

#neuroskyence

bit.ly/47MXYLH

10.11.2025 14:56 πŸ‘ 58 πŸ” 31 πŸ’¬ 1 πŸ“Œ 5
Preview
Whole-cortex in situ sequencing reveals input-dependent area identity - Nature BARseq interrogates the expression of 104 cell-type marker genes in 10.3 million cells over nine mouse forebrain hemispheres to reveal the role of peripheral inputs on cortical area development.

Still think brain regions don’t exist? That everything is everywhere? That cell types don’t matter and that everything is a dynamical phase portrait?

Wrong.

Interconnected brain modules exist at the level of fine grained transcriptomics. www.nature.com/articles/s41...

06.11.2025 12:39 πŸ‘ 69 πŸ” 20 πŸ’¬ 3 πŸ“Œ 0

Also need a new chapter on the cancer resistance mechanisms of the whale. πŸ“– πŸ‹ www.nature.com/articles/s41...

04.11.2025 02:59 πŸ‘ 8 πŸ” 1 πŸ’¬ 0 πŸ“Œ 0

Pre-stimulus pupil-linked arousal enhances initial stimulus availability and accelerates decay in iconic memory. https://www.biorxiv.org/content/10.1101/2025.10.31.685757v1

01.11.2025 12:15 πŸ‘ 1 πŸ” 1 πŸ’¬ 0 πŸ“Œ 0
Preview
Facial expressions in mice reveal latent cognitive variables and their neural correlates - Nature Neuroscience The face reveals more than just emotion. Cazettes, Reato and colleagues show that subtle facial movements reveal hidden cognitive states, reflecting the brain’s ongoing computations and offering a noninvasive window into unexpressed thoughts and decisions.

Characteristic facial features in mice simultaneously encode not only currently used decision variables but also independent and unexpressed ones, and these features partially originate from neural activity in the secondary motor cortex

@fannycazettes.bsky.social

www.nature.com/articles/s41...

03.10.2025 15:45 πŸ‘ 27 πŸ” 6 πŸ’¬ 1 πŸ“Œ 0
Preview
Ultra-high-density Neuropixels probes improve detection and identification in neuronal recordings To understand the neural basis of behavior, it is essential to sensitively and accurately measure neural activity at single-neuron and single-spike re…

Neuropixels Ultra!

By @steinmetzneuro.bsky.social, who led a big collaboration

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

30.09.2025 15:11 πŸ‘ 81 πŸ” 18 πŸ’¬ 0 πŸ“Œ 3

Learning induces persistent chromatin loops underlying robust gene expression during memory recall https://www.biorxiv.org/content/10.1101/2025.09.30.678146v1

30.09.2025 21:15 πŸ‘ 1 πŸ” 1 πŸ’¬ 0 πŸ“Œ 0
Preview
Huntington's disease successfully treated for first time One of the most devastating diseases finally has a treatment that can slow its progression and transform lives, tearful doctors tell BBC.

In the early years of my PhD I remember being asked if we'd ever treat or cure diseases like HIV or Huntington's. I remember saying maybe, but not knowing if I believed. Seeing research pay off as real impact on human lives is - incredible.

www.bbc.com/news/article...

www.bbc.com/news/article...

24.09.2025 12:41 πŸ‘ 148 πŸ” 41 πŸ’¬ 8 πŸ“Œ 4
Post image

You could have rather put the quotation marks in "arousal", given the last sentence of the paper.
Nice and impressive, but it's basically saying that there is a low-dimensional descriptor. Now, it could have been reversed having latent brain dynamics as a predictor, and pupillometry as the target

25.09.2025 14:24 πŸ‘ 2 πŸ” 1 πŸ’¬ 0 πŸ“Œ 0
Preview
Arousal as a universal embedding for spatiotemporal brain dynamics - Nature Reframing of arousal as a latent dynamical system can reconstruct multidimensional measurements of large-scale spatiotemporal brain dynamics on the timescale of seconds in mice.

Super proud of this collaboration with rockstar Ryan Raut - born out of playing in the sandbox in our last year of grad school! Multi-scale brain activity can be predicted from a simple measure of arousal like pupil diameter. Out with linear causality, in with dynamic systems to explain neurobiology

24.09.2025 21:52 πŸ‘ 85 πŸ” 25 πŸ’¬ 3 πŸ“Œ 4
Preview
Arousal as a universal embedding for spatiotemporal brain dynamics - Nature Reframing of arousal as a latent dynamical system can reconstruct multidimensional measurements of large-scale spatiotemporal brain dynamics on the timescale of seconds in mice.

This one looks intriguing. Arousal "embedding" whole-brain dynamics. 🀯

#neuroskyence #compneurosky

doi.org/10.1038/s415...

25.09.2025 06:03 πŸ‘ 49 πŸ” 17 πŸ’¬ 4 πŸ“Œ 1
Post image

It’s been a 24 hour traveling day from Portland to PΓ©cs, which includes 2 planes, a bus, a metro, a train & a car.

I am so excited to be at ERM2025 and to be selected as a speaker in the New Perspectives, Tools and Resources session!

21.09.2025 16:42 πŸ‘ 6 πŸ” 2 πŸ’¬ 0 πŸ“Œ 0
Preview
A new look at the pupil: more than just a gatekeeper | EMBL New research from the Rompani and Asari groups at EMBL Rome suggests that the pupil does more than just respond to light and internal states β€” it may actively shape our vision.

Our EMBL Press office has a nice little article on our recent paper on how the pupil's dynamics shape visual responses in the mouse and human.

A new look at the pupil: more than just a gatekeeper www.embl.org/news/science...

20.09.2025 09:34 πŸ‘ 4 πŸ” 1 πŸ’¬ 0 πŸ“Œ 0
Preview
Granularity of thalamic head direction cells Head direction signaling is fundamental for spatial orientation and navigation. The anterodorsal nucleus of the thalamus (ADn) contains a high density of head direction (HD) cells that process sensori...

*New preprint from the lab* – β€œGranularity of thalamic head direction cells”
doi.org/10.1101/2025...
πŸ§ πŸ“ˆ πŸ§ͺ
1/11

15.09.2025 15:13 πŸ‘ 26 πŸ” 10 πŸ’¬ 2 πŸ“Œ 2