Alireza Karami | علیرضا کرمی

Interesting PhD positions in AI & ML with the ELLIS–Max Planck AI Network. Potential opportunities for interdisciplinary projects bridging AI and neuroscience.

Variance partitioning is used to quantify the overlap of two models. Over the years, I have found that this can be a very confusing and misleading concept. So we finally we decided to write a short blog to explain why.
@martinhebart.bsky.social @gallantlab.org
diedrichsenlab.org/BrainDataSci...

I wanted to add some thoughts to this excellent blog post, not detailed, maybe wrong, maybe useful:
1. Unique variance is easy to interpret as a lower bound of what a variable explains (the upper bound being either what the variable explains alone or what the other variables cannot explain uniquely)

Dissecting LLMs with ML | Mike X Cohen | Substack Understand how large language models (LLMs) really work by applying machine learning (ML) methods to their internal activations. Each post explores how LLMs process text, isolate patterns, and generat...

Dear scientists of BlueSky: Wouldn't it be great to learn how large language models work by running experiments on them and analyzing the data? Yeah, you can do that, and I'll show you how.
mikexcohen.substack.com/s/ml-on-llms

I’m recruiting committee members for the Technical Program Committee at #CCN2026.

Please apply if you want to help make submission, review & selection of contributed work (Extended Abstracts & Proceedings) more useful for everyone! 🌐

Helps to have: programming/communications/editorial experience.

Thrilled that #CCN2026 will be hosted at NYU! Having experienced an amazing summer school there last year, I know how inspiring the city is. It’s heartbreaking, though, that many brilliant Iranian students and researchers may be excluded due to the new travel restrictions.

Come and check out our poster at #CCN2025, presented by @tlmnhut.bsky.social

Investigating action topography in visual cortex and deep artificial neural networks High-level visual cortex contains category-selective areas embedded within larger-scale topographic maps like animacy and real-world size. Here, we propose action as a key organizing factor shaping vi...

New preprint out! We propose that action is a key dimension shaping the topographic organization of object categories in lateral occipitotemporal cortex (LOTC)—and test whether standard and topographic neural networks capture this pattern. A thread:

www.biorxiv.org/content/10.1...

🧵 1/n

We’re very happy to share 3 highlights of our Brain and AI team for #CCN2025 's week:

1. 🏆1st place for the Algonauts competition: paper, thtread and code below

2.🗣Keynote: Language in the Brain: 2025.ccneuro.org/k-and-t-lang...

3. 🚀Tutorial: Scale your decoding pipeline in the notebook

2️⃣ A poster led by Samuel Debray, with @standehaene.bsky.social — another exciting project.

📄tinyurl.com/2zt6h8z7

Missed #CCN2025 this year, but still excited to share two works there!

1️⃣ From my PhD with @manpiazza.bsky.social — accepted in the CCN proceedings. My young collaborator @tlmnhut.bsky.social will be presenting it. It’s about numerosity representation in CNNs.

📄 tinyurl.com/yc2dyhm3

The Journal of Neuroscience: 45 (28)

Excited to share our JNeurosci paper- featuring on the cover!
The Representational Organization of Static and Dynamic Visual Features in the Human Cortex
w/ Jianxin Wang, and Stefano Anzellotti
Cover: www.jneurosci.org/content/45/28.cover-expansion
Paper: www.jneurosci.org/content/45/28/e1164242025

A computational linguistics colleague needed to join the Center for Mind-Brain Sciences - Italy's leading cognitive neuroscience research unit. @cimecunitrento.bsky.social is located in Trentino, famous for its high quality of life in Italy. www.cimec.unitn.it/en/71/langua...

Looking for a Master internship and interested in timing the interplay between timing and hearing?
🔽
brainthemind.com/openings/

Seeing the Mind, Educating the Brain

If you are in Paris on October 1-3 : we are organizing a fantastic cognitive neuroscience conference at Collège de France, on topics ranging from language to math, education and consciousness, with many of my favorite scientists !
Full program here:
www.unicog.org/seeing-the-m...

Distinct neural representational geometries of numerosity in early visual and association regions across visual streams - Communications Biology This study reveals distinct neural codes for visual numerosity across the brain, showing that numerosity is independently represented from early visual areas to higher-level regions along both dorsal ...

Huge thanks to the @cimecunitrento.bsky.social fMRI lab, and to Ben Harvey and Florent Meyniel for inspiring discussions. 🙏

And of course, to the participants who made this work possible.

Read the full paper here:
🔗 www.nature.com/articles/s42...

💬 Curious to discuss? My DMs are open!

Exciting avenues ahead:
✅ Developmental studies—do these ventral representations emerge early or with formal education?
✅ Cross-modal generalization—does the same geometry hold for other sensory modalities?
✅ Links to behavior—how do these neural patterns relate to numerical skills?

🌟 Our results challenge the idea that numerosity is mainly a dorsal stream function.

Ventral areas also show robust encoding—suggesting a more distributed system than @standehaene.bsky.social ’s Triple Code Model.

This raises new questions about symbolic and non-symbolic number processing.

• Early visual areas showed a linear number line.
• Parietal and ventral regions showed a curved manifold, separating extremes from middle numerosities.

Possible interp.:
🔸 Numerosity-tuned coding @jacobmpaul.bsky.social aul.bsky.social‬
🔸 Decision-variable encoding @summerfieldlab.bsky.social

🧠 We replicate Castaldi et al. (2019), showing numerosity is encoded independently of other visual features along the dorsal stream.

🔍 We extend this to the ventral stream, where number representations were often neglected, aligning with topographic maps reported by @sergedumoulin.net & Harvey.

Our approach built on Castaldi et al. (2019) and combined:
🔹 RSA—to quantify how numerical and non-numerical features are encoded
🔹 Multidimensional Scaling (MDS)—to visualize the neural geometry of numerosity representations

This allowed us to isolate number representations from visual confounds.

We used fMRI in 31 adults performing a numerosity estimation task.
Visual arrays varied in:
• Number of dots
• Item size
• Total area
• Density

✅ Critically controlled for visual confounds to isolate genuine number representations.

Why care about numerosity?
Numerosity perception is a core cognitive skill:
• Present early in development
• Shared across species
• Thought to scaffold formal math learning

Yet, where and how the brain encodes it independently of other visual features has remained debated.

New paper out in @commsbio.nature.com!
@elicastaldi.bsky.social, Evelyn Eger, @manpiazza.bsky.social

🔢 We reveal how the brain represents numerosity across the entire visual system, from early visual areas to high-level association cortices.

👇 A thread
🔗 www.nature.com/articles/s42...

🎉 New Chapter Published!

🧠 Title:
Neurobiological underpinnings of developmental dyscalculia
👥 Authors:
Paula A. Maldonado Moscoso, @alirezakr.bsky.social , @manpiazza.bsky.social

#Neuroscience #CognitiveNeuroscience #LearningDisabilities #Dyscalculia

New article out featuring CIMeC researchers Alireza Karami and Manuela Piazza, in collaboration with Elisa Castaldi and Evelyn Eger!

@alirezakr.bsky.social, @manpiazza.bsky.social

Read more 👇🏻
www.nature.com/articles/s42003-025-08395-z

Distinct neural representational geometries of numerosity in early visual and association regions across visual streams - Communications Biology This study reveals distinct neural codes for visual numerosity across the brain, showing that numerosity is independently represented from early visual areas to higher-level regions along both dorsal ...

This study reveals distinct neural codes for visual numerosity across the brain, showing that numerosity is independently represented from early visual areas to higher-level regions along both dorsal and ventral streams. www.nature.com/articles/s42...

🔗 Read more here:
👉 tinyurl.com/muw8aeta

🔍 What’s it about?
Developmental dyscalculia is a learning disorder where children struggle with math despite normal intelligence and typical development.
✨ Key Highlights:
Consistent evidence of structural and functional changes in the right parietal cortex & its connections to frontal regions.

📚 Published in:
Developmental Dyscalculia: From Brain Mechanisms to Educational Applications
Edited by Yarden Gliksman, Liane Kaufmann, and Avishai Henik