Fernando Pérez-García
@fepegar.com
Senior research machine learning engineer at Microsoft Research Health Futures. PhD in Medical Imaging. Open source & open access. Created TorchIO. He/him.
Pinned
Fernando Pérez-García
@fepegar.com
· Nov 22
I’ve created a starter pack! Follow us if you want to know more about what we’re up to at Microsoft Research Health Futures 🩻⚕️🏥
Get in touch if you think you or someone else should be added!
Get in touch if you think you or someone else should be added!
Reposted by Fernando Pérez-García
TLDR; The PSF has made the decision to put our community and our shared diversity, equity, and inclusion values ahead of seeking $1.5M in new revenue. Please read and share. pyfound.blogspot.com/2025/10/NSF-...
🧵
🧵
The official home of the Python Programming Language
www.python.org
October 27, 2025 at 2:47 PM
TLDR; The PSF has made the decision to put our community and our shared diversity, equity, and inclusion values ahead of seeking $1.5M in new revenue. Please read and share. pyfound.blogspot.com/2025/10/NSF-...
🧵
🧵
Reposted by Fernando Pérez-García
🩻Excited to share our latest preprint: “Data Scaling Laws for Radiology Foundation Models”
Foundation vision encoders like CLIP and DINOv2 have transformed general computer vision, but what happens when we scale them for medical imaging?
📄 Read the full preprint here: arxiv.org/abs/2509.12818
Foundation vision encoders like CLIP and DINOv2 have transformed general computer vision, but what happens when we scale them for medical imaging?
📄 Read the full preprint here: arxiv.org/abs/2509.12818
Data Scaling Laws for Radiology Foundation Models
Foundation vision encoders such as CLIP and DINOv2, trained on web-scale data, exhibit strong transfer performance across tasks and datasets. However, medical imaging foundation models remain constrai...
arxiv.org
September 23, 2025 at 8:34 AM
🩻Excited to share our latest preprint: “Data Scaling Laws for Radiology Foundation Models”
Foundation vision encoders like CLIP and DINOv2 have transformed general computer vision, but what happens when we scale them for medical imaging?
📄 Read the full preprint here: arxiv.org/abs/2509.12818
Foundation vision encoders like CLIP and DINOv2 have transformed general computer vision, but what happens when we scale them for medical imaging?
📄 Read the full preprint here: arxiv.org/abs/2509.12818
Reposted by Fernando Pérez-García
Recognition is no doubt significant but it will be a hollow gesture if the UK does not also seek to end Israel's genocide, illegal occupation, and system of apartheid against the Palestinian people.
🧵1/3
🧵1/3
September 22, 2025 at 12:04 PM
Recognition is no doubt significant but it will be a hollow gesture if the UK does not also seek to end Israel's genocide, illegal occupation, and system of apartheid against the Palestinian people.
🧵1/3
🧵1/3
Reposted by Fernando Pérez-García
Yeah man we should really fight back by staying on X
September 21, 2025 at 2:58 AM
Yeah man we should really fight back by staying on X
Having some fun with DINOv3 and PCA! Although I'm not happy my nose has such a low foreground probability :D
August 21, 2025 at 10:15 PM
Having some fun with DINOv3 and PCA! Although I'm not happy my nose has such a low foreground probability :D
Reposted by Fernando Pérez-García
New blog post: AI Allergy.
On my increasing disgust with the AI discourse, even though I still like the technical and philosophical. And how I wish I could be excited about AI again.
togelius.blogspot.com/2025/08/ai-a...
On my increasing disgust with the AI discourse, even though I still like the technical and philosophical. And how I wish I could be excited about AI again.
togelius.blogspot.com/2025/08/ai-a...
AI Allergy
I remember being excited about AI. I remember 20 years ago, being excited about neuroevolutionary methods for learning adaptive behaviors in...
togelius.blogspot.com
August 16, 2025 at 5:31 PM
Reposted by Fernando Pérez-García
Introducing DINOv3 🦕🦕🦕
A SotA-enabling vision foundation model, trained with pure self-supervised learning (SSL) at scale.
High quality dense features, combining unprecedented semantic and geometric scene understanding.
Three reasons why this matters👇
A SotA-enabling vision foundation model, trained with pure self-supervised learning (SSL) at scale.
High quality dense features, combining unprecedented semantic and geometric scene understanding.
Three reasons why this matters👇
August 14, 2025 at 6:52 PM
Introducing DINOv3 🦕🦕🦕
A SotA-enabling vision foundation model, trained with pure self-supervised learning (SSL) at scale.
High quality dense features, combining unprecedented semantic and geometric scene understanding.
Three reasons why this matters👇
A SotA-enabling vision foundation model, trained with pure self-supervised learning (SSL) at scale.
High quality dense features, combining unprecedented semantic and geometric scene understanding.
Three reasons why this matters👇
Reposted by Fernando Pérez-García
We’re not sure who needs to hear this, but ‘blueberry’ has two b’s.
August 8, 2025 at 8:38 PM
We’re not sure who needs to hear this, but ‘blueberry’ has two b’s.
Reposted by Fernando Pérez-García
Boiling here at home in Cyprus but I put the finishing touches a couple of days ago on this preprint: What Does 'Human-Centred AI' Mean? doi.org/10.48550/arX...
Wherein I analyse HCAI & demonstrate through 3 triplets my new tripartite definition of AI (Table 1) that properly centres the human. 1/n
Wherein I analyse HCAI & demonstrate through 3 triplets my new tripartite definition of AI (Table 1) that properly centres the human. 1/n
July 29, 2025 at 11:52 AM
Boiling here at home in Cyprus but I put the finishing touches a couple of days ago on this preprint: What Does 'Human-Centred AI' Mean? doi.org/10.48550/arX...
Wherein I analyse HCAI & demonstrate through 3 triplets my new tripartite definition of AI (Table 1) that properly centres the human. 1/n
Wherein I analyse HCAI & demonstrate through 3 triplets my new tripartite definition of AI (Table 1) that properly centres the human. 1/n
Reposted by Fernando Pérez-García
Scientists overwhelmingly recognize the value of sharing null results, but rarely publish them in the research literature
go.nature.com/450KElr
go.nature.com/450KElr
Researchers value null results, but struggle to publish them
Survey finds that fear of reputational harm and a lack of support and publication platforms are among respondents’ key concerns.
go.nature.com
July 23, 2025 at 12:22 PM
Scientists overwhelmingly recognize the value of sharing null results, but rarely publish them in the research literature
go.nature.com/450KElr
go.nature.com/450KElr
Reposted by Fernando Pérez-García
i love it when people make PRs really easy to review ❤️
feels really bad when there's a PR that's been open for a while, but i know i can't easily do a good job of reviewing it, so it's constantly "later"
feels really bad when there's a PR that's been open for a while, but i know i can't easily do a good job of reviewing it, so it's constantly "later"
July 23, 2025 at 11:32 AM
i love it when people make PRs really easy to review ❤️
feels really bad when there's a PR that's been open for a while, but i know i can't easily do a good job of reviewing it, so it's constantly "later"
feels really bad when there's a PR that's been open for a while, but i know i can't easily do a good job of reviewing it, so it's constantly "later"
Reposted by Fernando Pérez-García
New 2 year position as an Imaging Data Infrastructure Specialist
Located a the Hunter Medical Research Institute (HMRI) & the University of Newcastle (UoN), in partnership with the National Imaging Facility (NIF),
Apply here: www.seek.com.au/job/85537666
Located a the Hunter Medical Research Institute (HMRI) & the University of Newcastle (UoN), in partnership with the National Imaging Facility (NIF),
Apply here: www.seek.com.au/job/85537666
July 21, 2025 at 4:01 AM
New 2 year position as an Imaging Data Infrastructure Specialist
Located a the Hunter Medical Research Institute (HMRI) & the University of Newcastle (UoN), in partnership with the National Imaging Facility (NIF),
Apply here: www.seek.com.au/job/85537666
Located a the Hunter Medical Research Institute (HMRI) & the University of Newcastle (UoN), in partnership with the National Imaging Facility (NIF),
Apply here: www.seek.com.au/job/85537666
Reposted by Fernando Pérez-García
A #computervision researcher at #ICML2025
July 19, 2025 at 4:49 PM
A #computervision researcher at #ICML2025
Reposted by Fernando Pérez-García
New work from my team! arxiv.org/abs/2507.12950
Intersecting mechanistic interpretability and health AI 😎
We trained and interpreted sparse autoencoders on MAIRA-2, our radiology MLLM. We found a range of human-interpretable radiology reporting concepts, but also many uninterpretable SAE features.
Intersecting mechanistic interpretability and health AI 😎
We trained and interpreted sparse autoencoders on MAIRA-2, our radiology MLLM. We found a range of human-interpretable radiology reporting concepts, but also many uninterpretable SAE features.
Insights into a radiology-specialised multimodal large language model with sparse autoencoders
Interpretability can improve the safety, transparency and trust of AI models, which is especially important in healthcare applications where decisions often carry significant consequences. Mechanistic...
arxiv.org
July 18, 2025 at 9:30 AM
New work from my team! arxiv.org/abs/2507.12950
Intersecting mechanistic interpretability and health AI 😎
We trained and interpreted sparse autoencoders on MAIRA-2, our radiology MLLM. We found a range of human-interpretable radiology reporting concepts, but also many uninterpretable SAE features.
Intersecting mechanistic interpretability and health AI 😎
We trained and interpreted sparse autoencoders on MAIRA-2, our radiology MLLM. We found a range of human-interpretable radiology reporting concepts, but also many uninterpretable SAE features.
Reposted by Fernando Pérez-García
Just published in JOSS: 'Lighter: Configuration-Driven Deep Learning' https://doi.org/10.21105/joss.08101
July 11, 2025 at 3:52 PM
Just published in JOSS: 'Lighter: Configuration-Driven Deep Learning' https://doi.org/10.21105/joss.08101
Reposted by Fernando Pérez-García
The world’s first multimodal, bilingual radiology dataset could reshape the way radiologists and AI systems make sense of X-rays. PadChest-GR, developed by the University of Alicante with Microsoft Research, has the potential to advance research across the field for years to come. msft.it/6013SLDYZ
June 26, 2025 at 4:06 PM
The world’s first multimodal, bilingual radiology dataset could reshape the way radiologists and AI systems make sense of X-rays. PadChest-GR, developed by the University of Alicante with Microsoft Research, has the potential to advance research across the field for years to come. msft.it/6013SLDYZ
Reposted by Fernando Pérez-García
📣 Our new PadChest-GR benchmark in collaboration with @ua.es and MedBravo is published in @ai.nejm.org: ai.nejm.org/doi/full/10....!
📰 Also check out our blog post for why we're so excited about it: www.microsoft.com/en-us/resear...
💾 The dataset can be downloaded here: bimcv.cipf.es/bimcv-projec...
📰 Also check out our blog post for why we're so excited about it: www.microsoft.com/en-us/resear...
💾 The dataset can be downloaded here: bimcv.cipf.es/bimcv-projec...
June 26, 2025 at 5:01 PM
📣 Our new PadChest-GR benchmark in collaboration with @ua.es and MedBravo is published in @ai.nejm.org: ai.nejm.org/doi/full/10....!
📰 Also check out our blog post for why we're so excited about it: www.microsoft.com/en-us/resear...
💾 The dataset can be downloaded here: bimcv.cipf.es/bimcv-projec...
📰 Also check out our blog post for why we're so excited about it: www.microsoft.com/en-us/resear...
💾 The dataset can be downloaded here: bimcv.cipf.es/bimcv-projec...
Reposted by Fernando Pérez-García
A novel dataset designed to train and evaluate grounded report generation models from chest x-ray images includes comprehensive sentence-level bounding-box annotations for all clinically relevant findings in each image. Learn more: nejm.ai/44rk1a1
@dccastr0.bsky.social #AI #MedSky #MLSky
@dccastr0.bsky.social #AI #MedSky #MLSky
July 3, 2025 at 1:15 PM
A novel dataset designed to train and evaluate grounded report generation models from chest x-ray images includes comprehensive sentence-level bounding-box annotations for all clinically relevant findings in each image. Learn more: nejm.ai/44rk1a1
@dccastr0.bsky.social #AI #MedSky #MLSky
@dccastr0.bsky.social #AI #MedSky #MLSky
I am disappointed in the AI discourse steveklabnik.com/writing/i-am...
I am disappointed in the AI discourse
steveklabnik.com
June 29, 2025 at 12:56 PM
Reposted by Fernando Pérez-García
codingAssistantsAreJustCasinosForProgrammers
June 27, 2025 at 6:00 AM
codingAssistantsAreJustCasinosForProgrammers
Congratulations/enhorabuena to @dccastr0.bsky.social and the rest of the team 👏
One important step towards usability of AI in real-world radiology applications.
One important step towards usability of AI in real-world radiology applications.
The world’s first multimodal, bilingual radiology dataset could reshape the way radiologists and AI systems make sense of X-rays. PadChest-GR, developed by the University of Alicante with Microsoft Research, has the potential to advance research across the field for years to come. msft.it/6013SLDYZ
June 26, 2025 at 6:03 PM
Congratulations/enhorabuena to @dccastr0.bsky.social and the rest of the team 👏
One important step towards usability of AI in real-world radiology applications.
One important step towards usability of AI in real-world radiology applications.
Reposted by Fernando Pérez-García
Demystifying conda packages!
How do you make compiled software with hardcoded paths work anywhere?
255-character placeholders, $ORIGIN rpaths, and platform-specific binary patching.
We explain how Conda packages achieve relocatability:
prefix.dev/blog/what-i...
255-character placeholders, $ORIGIN rpaths, and platform-specific binary patching.
We explain how Conda packages achieve relocatability:
prefix.dev/blog/what-i...
What is a Conda package, actually?
At its core, a conda package really is just a "glorified" tarball—a compressed archive of files with some metadata attached.
prefix.dev
June 12, 2025 at 6:14 AM
Demystifying conda packages!
Reposted by Fernando Pérez-García
A new use of the asterisk in the paper author list for credit assignment
May 18, 2025 at 6:23 PM
A new use of the asterisk in the paper author list for credit assignment
You can think LLMs are aren’t intelligent, are built in exploitative ways, aren’t useful, overhyped, but if you don’t think they’re a wildly interesting scientific artifact I don’t know what to tell you
May 18, 2025 at 1:44 PM