Matteo Carandini
@carandinilab.net
Neuroscientist at University College London (www.ucl.ac.uk/cortexlab). Opinions my own.
Pinned
Matteo Carandini
@carandinilab.net
· Jul 7
New by Agnès Landemard (@agnesland.bsky.social) & co
Brainwide blood volume reflects opposing neural populations
Brainwide fluctuations in blood volume arise from two populations with opposite relation to brain state and distinct relationships to blood supply
www.biorxiv.org/content/10.1...
Brainwide blood volume reflects opposing neural populations
Brainwide fluctuations in blood volume arise from two populations with opposite relation to brain state and distinct relationships to blood supply
www.biorxiv.org/content/10.1...
This 2017 preprint by @suryaganguli.bsky.social is foundational: the dimension of a neural code can't be > that of the experiment used to measure it. Yet it is not very widely known. Maybe because people don't like the message? And because it's "unpublished"? (After writing it, Gao left for SpaceX)
Low-dimensional activity for simple tasks is a clear prediction from basic coding arguments
A theory of multineuronal dimensionality, dynamics and measurement
In many experiments, neuroscientists tightly control behavior, record many trials, and obtain trial-averaged firing rates from hundreds of neurons in circuits containing billions of behaviorally relev...
www.biorxiv.org
November 24, 2025 at 9:16 AM
This 2017 preprint by @suryaganguli.bsky.social is foundational: the dimension of a neural code can't be > that of the experiment used to measure it. Yet it is not very widely known. Maybe because people don't like the message? And because it's "unpublished"? (After writing it, Gao left for SpaceX)
Reposted by Matteo Carandini
Hearing a rumor that NIH/HHS may now plan to retaliate against other Bethesda Declaration signers.
I am one, but it’s not about me: we have been speaking out because of the ongoing damage to US medical research, cancer and Alzheimer’s research.
As @jenna-m-norton.bsky.social put well: 1/
I am one, but it’s not about me: we have been speaking out because of the ongoing damage to US medical research, cancer and Alzheimer’s research.
As @jenna-m-norton.bsky.social put well: 1/
November 14, 2025 at 4:17 PM
Hearing a rumor that NIH/HHS may now plan to retaliate against other Bethesda Declaration signers.
I am one, but it’s not about me: we have been speaking out because of the ongoing damage to US medical research, cancer and Alzheimer’s research.
As @jenna-m-norton.bsky.social put well: 1/
I am one, but it’s not about me: we have been speaking out because of the ongoing damage to US medical research, cancer and Alzheimer’s research.
As @jenna-m-norton.bsky.social put well: 1/
Reposted by Matteo Carandini
It is a small trip off the NIH campus in order to post this, but I have a bike and the weather is excellent!
November 14, 2025 at 4:33 PM
It is a small trip off the NIH campus in order to post this, but I have a bike and the weather is excellent!
Reposted by Matteo Carandini
Image screening is going to fail. We need audit trails for data provenance.
I tried an even harder example on Gemini Pro image generation and this is quite scary/amazing. I asked for a microscopy image of around 20 HeLa cells, GFP tagged 20% nuclear, 10% membrane, +1 nuclear staining, + overlap. Image below and prompt in the following post.
November 22, 2025 at 4:21 PM
Image screening is going to fail. We need audit trails for data provenance.
Reposted by Matteo Carandini
Sometimes it feels like the LLM is a Prima Donna, and the prompt is a motivational pep talk putting it in just the right frame of mind to get peak performance.
November 21, 2025 at 3:09 PM
Sometimes it feels like the LLM is a Prima Donna, and the prompt is a motivational pep talk putting it in just the right frame of mind to get peak performance.
Reposted by Matteo Carandini
🚨 Big news for the neuroscience community!
The Call for Abstracts, Travel Grants and Registrations for #FENS2026 is officially open! 🧠
👉 Register and submit your work today: https://loom.ly/usQR0ao
The Call for Abstracts, Travel Grants and Registrations for #FENS2026 is officially open! 🧠
👉 Register and submit your work today: https://loom.ly/usQR0ao
November 14, 2025 at 3:00 PM
🚨 Big news for the neuroscience community!
The Call for Abstracts, Travel Grants and Registrations for #FENS2026 is officially open! 🧠
👉 Register and submit your work today: https://loom.ly/usQR0ao
The Call for Abstracts, Travel Grants and Registrations for #FENS2026 is officially open! 🧠
👉 Register and submit your work today: https://loom.ly/usQR0ao
Reposted by Matteo Carandini
This is an excellent blueprint on a very fascinating use of AI scientist! And the results and super cool and interesting! 🤩
I have been asked this when talking about our work on using powerlaws to study representation quality in deep neural networks, glad to have a more concrete answer now! 😃
I have been asked this when talking about our work on using powerlaws to study representation quality in deep neural networks, glad to have a more concrete answer now! 😃
1. New preprint resolving a conundrum in systems neuroscience with an AI scientist, and humans Reilly Tilbury, Dabin Kwon, @haydari.bsky.social, @jacobmratliff.bsky.social, @bio-emergent.bsky.social, @carandinilab.net, @kevinjmiller.bsky.social, @neurokim.bsky.social
www.biorxiv.org/content/10.1...
www.biorxiv.org/content/10.1...
Characterizing neuronal population geometry with AI equation discovery
The visual cortex contains millions of neurons, whose combined activity forms a population code representing visual stimuli. There is, however, a discrepancy between our understanding of this code at ...
www.biorxiv.org
November 16, 2025 at 10:29 PM
This is an excellent blueprint on a very fascinating use of AI scientist! And the results and super cool and interesting! 🤩
I have been asked this when talking about our work on using powerlaws to study representation quality in deep neural networks, glad to have a more concrete answer now! 😃
I have been asked this when talking about our work on using powerlaws to study representation quality in deep neural networks, glad to have a more concrete answer now! 😃
Reposted by Matteo Carandini
I've been super excited about @kevinjmiller.bsky.social and @neurokim.bsky.social's AI scientist project since I heard about it
delighted to see them collaborating here with the awesome @kenneth-harris.bsky.social and lab to create full-on new science and new scientific insights.
It's the future!
delighted to see them collaborating here with the awesome @kenneth-harris.bsky.social and lab to create full-on new science and new scientific insights.
It's the future!
1. New preprint resolving a conundrum in systems neuroscience with an AI scientist, and humans Reilly Tilbury, Dabin Kwon, @haydari.bsky.social, @jacobmratliff.bsky.social, @bio-emergent.bsky.social, @carandinilab.net, @kevinjmiller.bsky.social, @neurokim.bsky.social
www.biorxiv.org/content/10.1...
www.biorxiv.org/content/10.1...
Characterizing neuronal population geometry with AI equation discovery
The visual cortex contains millions of neurons, whose combined activity forms a population code representing visual stimuli. There is, however, a discrepancy between our understanding of this code at ...
www.biorxiv.org
November 16, 2025 at 3:13 AM
I've been super excited about @kevinjmiller.bsky.social and @neurokim.bsky.social's AI scientist project since I heard about it
delighted to see them collaborating here with the awesome @kenneth-harris.bsky.social and lab to create full-on new science and new scientific insights.
It's the future!
delighted to see them collaborating here with the awesome @kenneth-harris.bsky.social and lab to create full-on new science and new scientific insights.
It's the future!
Reposted by Matteo Carandini
Now that the Sept and Oct data are (belatedly) in, it looks like 2025 will be the second warmest year in the record (~80% probability). The last three years are in a class of their own.
November 15, 2025 at 4:54 PM
Now that the Sept and Oct data are (belatedly) in, it looks like 2025 will be the second warmest year in the record (~80% probability). The last three years are in a class of their own.
Reposted by Matteo Carandini
1. Overshoot 1.5°C unavoidable, but returning below is possible
Every five years without substantial CO2 emission reductions locks in an additional temperature increase of about 0.1°C, and implies an additional need of around 200 billion tonnes CO2 of negative emissions to return to 1.5°C.
Every five years without substantial CO2 emission reductions locks in an additional temperature increase of about 0.1°C, and implies an additional need of around 200 billion tonnes CO2 of negative emissions to return to 1.5°C.
Búzios Scientific Statement
100 climate researchers, coordinated by @pik-potsdam.bsky.social & @iiasa.ac.at, released a joint statement to inform the #COP30 negotiations and provide guidance on difficult but doable global efforts designed to limit dangerous overshoot of 1.5°C as much as possible.
100 climate researchers, coordinated by @pik-potsdam.bsky.social & @iiasa.ac.at, released a joint statement to inform the #COP30 negotiations and provide guidance on difficult but doable global efforts designed to limit dangerous overshoot of 1.5°C as much as possible.
Búzios Scientific Statement.pdf
drive.google.com
November 15, 2025 at 9:01 PM
1. Overshoot 1.5°C unavoidable, but returning below is possible
Every five years without substantial CO2 emission reductions locks in an additional temperature increase of about 0.1°C, and implies an additional need of around 200 billion tonnes CO2 of negative emissions to return to 1.5°C.
Every five years without substantial CO2 emission reductions locks in an additional temperature increase of about 0.1°C, and implies an additional need of around 200 billion tonnes CO2 of negative emissions to return to 1.5°C.
Reposted by Matteo Carandini
6. The AI scientist took 45 minutes and $8.25 in LLM tokens to find a new tuning equation that fits the data better, and predicts the population code’s high-dimensional structure – even though we had only tasked it to model single-cell tuning.
November 14, 2025 at 6:07 PM
6. The AI scientist took 45 minutes and $8.25 in LLM tokens to find a new tuning equation that fits the data better, and predicts the population code’s high-dimensional structure – even though we had only tasked it to model single-cell tuning.
Reposted by Matteo Carandini
1. New preprint resolving a conundrum in systems neuroscience with an AI scientist, and humans Reilly Tilbury, Dabin Kwon, @haydari.bsky.social, @jacobmratliff.bsky.social, @bio-emergent.bsky.social, @carandinilab.net, @kevinjmiller.bsky.social, @neurokim.bsky.social
www.biorxiv.org/content/10.1...
www.biorxiv.org/content/10.1...
Characterizing neuronal population geometry with AI equation discovery
The visual cortex contains millions of neurons, whose combined activity forms a population code representing visual stimuli. There is, however, a discrepancy between our understanding of this code at ...
www.biorxiv.org
November 14, 2025 at 6:07 PM
1. New preprint resolving a conundrum in systems neuroscience with an AI scientist, and humans Reilly Tilbury, Dabin Kwon, @haydari.bsky.social, @jacobmratliff.bsky.social, @bio-emergent.bsky.social, @carandinilab.net, @kevinjmiller.bsky.social, @neurokim.bsky.social
www.biorxiv.org/content/10.1...
www.biorxiv.org/content/10.1...
Reposted by Matteo Carandini
Don't skip this. It's masterfully done, and many of the key points are GENERAL and apply to many other fields.
"there can sometimes just be lots of smoke"
I discuss the general case here: labrigger.com/blog/2025/11...
"there can sometimes just be lots of smoke"
I discuss the general case here: labrigger.com/blog/2025/11...
November 13, 2025 at 8:07 PM
Don't skip this. It's masterfully done, and many of the key points are GENERAL and apply to many other fields.
"there can sometimes just be lots of smoke"
I discuss the general case here: labrigger.com/blog/2025/11...
"there can sometimes just be lots of smoke"
I discuss the general case here: labrigger.com/blog/2025/11...
Our lab is looking for a postdoc! We have interesting projects and cutting-edge techniques such as Neuropixels Opto, Light Beads Microscopy and more. We would be delighted to receive your application. Deadline is 25 November 2025. More info here:
www.ucl.ac.uk/cortexlab/po...
www.ucl.ac.uk/cortexlab/po...
November 12, 2025 at 3:41 PM
Our lab is looking for a postdoc! We have interesting projects and cutting-edge techniques such as Neuropixels Opto, Light Beads Microscopy and more. We would be delighted to receive your application. Deadline is 25 November 2025. More info here:
www.ucl.ac.uk/cortexlab/po...
www.ucl.ac.uk/cortexlab/po...
Reposted by Matteo Carandini
The end of the internal combustion engine?
Sales of petrol and diesel cars in China peaked in 2017 and are now falling rapidly.
The shift to electric and low-carbon transport is well underway—and EV sales are rising fast.
Sales of petrol and diesel cars in China peaked in 2017 and are now falling rapidly.
The shift to electric and low-carbon transport is well underway—and EV sales are rising fast.
November 12, 2025 at 3:36 PM
The end of the internal combustion engine?
Sales of petrol and diesel cars in China peaked in 2017 and are now falling rapidly.
The shift to electric and low-carbon transport is well underway—and EV sales are rising fast.
Sales of petrol and diesel cars in China peaked in 2017 and are now falling rapidly.
The shift to electric and low-carbon transport is well underway—and EV sales are rising fast.
Reposted by Matteo Carandini
All talks of the 2025 UCL #Neuropixels course are now available: www.youtube.com/playlist?lis...
For more information on the course see the website: www.ucl.ac.uk/neuropixels/...
And for those at #SfN, come say hi at the #Neuropixels booth (Number 3731) or at my Poster 👋.
Happy Neuropixeling!
For more information on the course see the website: www.ucl.ac.uk/neuropixels/...
And for those at #SfN, come say hi at the #Neuropixels booth (Number 3731) or at my Poster 👋.
Happy Neuropixeling!
2025 UCL Neuropixels Course - YouTube
Lectures at the 2025 UCL Neuropixels course (https://www.ucl.ac.uk/neuropixels/training/2025-neuropixels-course)
www.youtube.com
November 10, 2025 at 1:48 PM
All talks of the 2025 UCL #Neuropixels course are now available: www.youtube.com/playlist?lis...
For more information on the course see the website: www.ucl.ac.uk/neuropixels/...
And for those at #SfN, come say hi at the #Neuropixels booth (Number 3731) or at my Poster 👋.
Happy Neuropixeling!
For more information on the course see the website: www.ucl.ac.uk/neuropixels/...
And for those at #SfN, come say hi at the #Neuropixels booth (Number 3731) or at my Poster 👋.
Happy Neuropixeling!
Reposted by Matteo Carandini
On the event of James Watson's death, I highly recommend this 2023 commentary from @matthewcobb.bsky.social and Nathaniel Comfort with crucial new insights into the discovery of the double helix. (And also check out Cobb's brand new biography of Francis Crick) www.nature.com/articles/d41...
What Rosalind Franklin truly contributed to the discovery of DNA’s structure
Franklin was no victim in how the DNA double helix was solved. An overlooked letter and an unpublished news article, both written in 1953, reveal that she was an equal player.
www.nature.com
November 7, 2025 at 9:25 PM
On the event of James Watson's death, I highly recommend this 2023 commentary from @matthewcobb.bsky.social and Nathaniel Comfort with crucial new insights into the discovery of the double helix. (And also check out Cobb's brand new biography of Francis Crick) www.nature.com/articles/d41...
Reposted by Matteo Carandini
Imagine if someone recursively trained LLMs to write, review, and edit papers. What would scientific papers look like after ~1000 epochs? How bad would the model collapse be? How much new input would be needed to avoid it?
November 7, 2025 at 5:44 PM
Imagine if someone recursively trained LLMs to write, review, and edit papers. What would scientific papers look like after ~1000 epochs? How bad would the model collapse be? How much new input would be needed to avoid it?
Reposted by Matteo Carandini
Sigh. NIH normally sends several hundred scientists to the SFN annual meeting to learn, exchange info, come up with new ideas, and advance science. (The exchange of ideas is the very core of the scientific enterprise.)
This year, no one from NIH will attend due to the gov't implosion.
This year, no one from NIH will attend due to the gov't implosion.
Enjoy a broad range of lectures and network with experts in the field at #SfN25! 🧠
Gain valuable insights and tools to enhance your research and career.
Browse through to lectures and build your itinerary in the 2025 Neuroscience Meeting Planner
🔗 vist.ly/4de72
#neurosky #neuroskyence
Gain valuable insights and tools to enhance your research and career.
Browse through to lectures and build your itinerary in the 2025 Neuroscience Meeting Planner
🔗 vist.ly/4de72
#neurosky #neuroskyence
November 7, 2025 at 10:53 PM
Sigh. NIH normally sends several hundred scientists to the SFN annual meeting to learn, exchange info, come up with new ideas, and advance science. (The exchange of ideas is the very core of the scientific enterprise.)
This year, no one from NIH will attend due to the gov't implosion.
This year, no one from NIH will attend due to the gov't implosion.
Reposted by Matteo Carandini
Calling aspiring group leaders! We've just opened applications for group leader fellowship sponsorship to start a lab in UCL NPP. Deadline 5 Jan. @uclnpp.bsky.social tinyurl.com/3h33dvev
UCL – University College London
UCL is consistently ranked as one of the top ten universities in the world (QS World University Rankings 2010-2022) and is No.2 in the UK for research power (Research Excellence Framework 2021).
tinyurl.com
November 6, 2025 at 10:06 AM
Calling aspiring group leaders! We've just opened applications for group leader fellowship sponsorship to start a lab in UCL NPP. Deadline 5 Jan. @uclnpp.bsky.social tinyurl.com/3h33dvev
Together with Musk's$1 tn pay deal and nVidia's $5 tn valuation, this will define our absurd era.
ig.ft.com/saudi-neom-l...
Oh, and "Fifteen members of the Huwaitat tribe who protested against their eviction were sent to prison, some for up to 50 years, and three others were sentenced to death"
ig.ft.com/saudi-neom-l...
Oh, and "Fifteen members of the Huwaitat tribe who protested against their eviction were sent to prison, some for up to 50 years, and three others were sentenced to death"
End of The Line: how Saudi Arabia’s Neom dream unravelled
Mohammed bin Salman’s utopian city was undone by the laws of physics and finance
ig.ft.com
November 6, 2025 at 10:15 PM
Together with Musk's$1 tn pay deal and nVidia's $5 tn valuation, this will define our absurd era.
ig.ft.com/saudi-neom-l...
Oh, and "Fifteen members of the Huwaitat tribe who protested against their eviction were sent to prison, some for up to 50 years, and three others were sentenced to death"
ig.ft.com/saudi-neom-l...
Oh, and "Fifteen members of the Huwaitat tribe who protested against their eviction were sent to prison, some for up to 50 years, and three others were sentenced to death"
Reposted by Matteo Carandini
🗣️ Edward Chang | Hertie Foundation Plenary Lecture
On 10 July at #FENS2026, Prof. Edward Chang (UCSF) will take us into the neural basis of speech, movement, emotion, and learning.
Meet all the #FENS2026 speakers 👉 https://loom.ly/Q_m95o8
On 10 July at #FENS2026, Prof. Edward Chang (UCSF) will take us into the neural basis of speech, movement, emotion, and learning.
Meet all the #FENS2026 speakers 👉 https://loom.ly/Q_m95o8
November 5, 2025 at 8:00 AM
🗣️ Edward Chang | Hertie Foundation Plenary Lecture
On 10 July at #FENS2026, Prof. Edward Chang (UCSF) will take us into the neural basis of speech, movement, emotion, and learning.
Meet all the #FENS2026 speakers 👉 https://loom.ly/Q_m95o8
On 10 July at #FENS2026, Prof. Edward Chang (UCSF) will take us into the neural basis of speech, movement, emotion, and learning.
Meet all the #FENS2026 speakers 👉 https://loom.ly/Q_m95o8
Reposted by Matteo Carandini
Choose France for Science!
Moved my lab to France 4 years ago — one of my best decisions.
Good life ✅
Strong talent pool ✅
Decent funding ✅
Byzantine bureaucracy ✅ (not much worse than UK/US, often better than Southern/Eastern EU)
Check out the call & reach out if interested
tinyurl.com/5n6u49uv
Moved my lab to France 4 years ago — one of my best decisions.
Good life ✅
Strong talent pool ✅
Decent funding ✅
Byzantine bureaucracy ✅ (not much worse than UK/US, often better than Southern/Eastern EU)
Check out the call & reach out if interested
tinyurl.com/5n6u49uv
Call for proposals details
anr.fr
November 4, 2025 at 8:32 PM
Choose France for Science!
Moved my lab to France 4 years ago — one of my best decisions.
Good life ✅
Strong talent pool ✅
Decent funding ✅
Byzantine bureaucracy ✅ (not much worse than UK/US, often better than Southern/Eastern EU)
Check out the call & reach out if interested
tinyurl.com/5n6u49uv
Moved my lab to France 4 years ago — one of my best decisions.
Good life ✅
Strong talent pool ✅
Decent funding ✅
Byzantine bureaucracy ✅ (not much worse than UK/US, often better than Southern/Eastern EU)
Check out the call & reach out if interested
tinyurl.com/5n6u49uv
Europe's MSCA postdoctoral fellowships are now attracting many researchers who would have gone to the US. So the payline has dropped below 10% (~1,600 fellowships to 17k applicants). Super tough situation. One more reason to double the next Horizon budget, as proposed by the EU Commission.
Applications for MCSA are up 64%! 50% now come from outside the EU. Trump effect?
The European Research market just got a hell of a lot more crowded.
EU should turn on the tap
marie-sklodowska-curie-actions.ec.europa.eu/news/msca-po...
The European Research market just got a hell of a lot more crowded.
EU should turn on the tap
marie-sklodowska-curie-actions.ec.europa.eu/news/msca-po...
MSCA Postdoctoral Fellowships 2025 receives record number of 17,058 proposals
This year’s annual MSCA Postdoctoral Fellowships call has attracted considerable interest from the research community with 17,058 proposals submitted.
marie-sklodowska-curie-actions.ec.europa.eu
November 3, 2025 at 8:24 AM
Europe's MSCA postdoctoral fellowships are now attracting many researchers who would have gone to the US. So the payline has dropped below 10% (~1,600 fellowships to 17k applicants). Super tough situation. One more reason to double the next Horizon budget, as proposed by the EU Commission.
Reposted by Matteo Carandini
Many animal studies promise interventions for early brain injury after subarachnoid hemorrhage, but treatments remain elusive. A possible reason? 40% of articles two data sleuths examined had evidence of duplicated/manipulated images. My latest:
www.thetransmitter.org/publishing/i...
www.thetransmitter.org/publishing/i...
Image integrity issues create new headache for subarachnoid hemorrhage research
First-time sleuths found potentially problematic images in hundreds of papers about early brain injury after subarachnoid hemorrhage.
www.thetransmitter.org
October 30, 2025 at 6:23 PM
Many animal studies promise interventions for early brain injury after subarachnoid hemorrhage, but treatments remain elusive. A possible reason? 40% of articles two data sleuths examined had evidence of duplicated/manipulated images. My latest:
www.thetransmitter.org/publishing/i...
www.thetransmitter.org/publishing/i...