Roman Pogodin Arna Ghosh Blake Richards Mila - Institut québécois d'IA I’m thrilled to announce I’ll be joining University of Leeds School of Computing to work on resource efficient AI and theories of biological intelligence! If you’re interested in work like this, I’ll be recruiting PhD students and postdocs. Please reach out or forward this on! 12/12

Lab’s latest out at PLOS Comp Biol, led by Carrisa V Cocuzza, PhD (she/her): “Distributed network flows generate localized category selectivity in human visual cortex”. This one changed how I think the brain works! Even "localized" functions are likely generated by distributed processes [1/N]

How does the brain represent the structure of visual scenes independent of their contents? In “Zero-shot counting with a dual-stream neural network model” (Neuron), we propose a model based on the parallel pathways of the primate visual system. shorturl.at/Zp5EA

Neuroscience is Marrist-Dominated in Sebastian and Jonathan Pillow s world. Yet I haven’t met anyone in practice breaking up a problem using Marrs framework ever. I can think of max 4 examples where people even tried to historically. Next we will announce lemarckianism is dead.

No, AI Has Not Killed Marr's Three Levels, It’s Merely Maimed Them (Consider me triggered. 😂🤣) The rapid advancement of artificial intelligence, particularly in neural networks and deep learning, has led some to question the relevance of classical frameworks like David Marr's

In light of recent news from across our southern border, I’m resharing this opportunity for anyone interested in spending the next couple of years up north — and meanwhile, doing a postdoc in NeuroAI in Montreal.

Two things going on here: 1) Text is special. Language has evolved to convey things concisely. We choose each word carefully to guide the listener toward some intended mental state; each word has high "counterfactual potency," in that most other choices would convey something

Prediction might still work for self-supervised learning in vision, but figuring out what to predict isn’t trivial. And scaling up alone won’t solve that problem.

CCN is now accepting full-length papers. An exciting development that provides a dedicated publication venue for works at the intersection of cognitive neuroscience and AI.

Félicitations aux lauréats du premier concours de communication scientifique Science éclair qui s'est tenu hier à Mila! 🥇 Dounia Shaaban Kabakibo 🥈 Arna Ghosh (Arna Ghosh) 🥉 Abdelrahman Zayed (Abdelrahman Zayed) Le concours a été conçu afin de souligner l'excellence académique de

She’s kinda wrong, because you could move to the other site and get the premium experience for free.

An example of biased media these days, dehumanizing one side: Iran killing civilians (which of course is tragic) vs Israel hitting military sites. No, Iranian civilians have been killed in every Israeli attack since Friday (in hundreds so far). Don't rebrand it. #StopTheWar

Let alone how ~400,000 people could be evacuated in less than an hour, or to where … We witnessed what they did to Gaza, and now we’re seeing and feeling the same strategies playing out in the city we grew up in.

It takes weeks if not months to evacuate Tehran.

A city with a population of 15 million, and all roads leading out are blocked. This is simply a murder threat.

I was just blocked by Peyman Milanfar 😂 after pointing out a common misconception: He claims "KL divergence was a definition not a derivation" This is incorrect. KL is a deep concept that can be derived from first principles 👉[1/5]🧵