Curious to know more? Read @philipcball.bsky.socialpress.uchicago.edu/ucp/books/bo...

How Life Works

“Bold and intriguing.”—Wall Street Journal • “Penetrating. . . . Provocative and profound.”—Publishers Weekly (starred review) • “Offers plenty of food for thought.”—Kirkus Reviews (starred review)  “...

This past week, I had a fantastic time attending NAISys at @cshlaboratory.bsky.social - a NeuroAI conference. It was amazing to learn about various topics like genomic coding, neuromorphic hardware, and in-context learning at the home to eight Nobel Prize winners in Physiology or Medicine! 1/N

Evidence that entorhinal cortex and prefrontal cortex (in mice, sorry) work together to encode outcomes in associative learning: www.nature.com/articles/s41... Studying this kind of outcome monitoring is going to be critical to understand the losses at play in the brain! 🧠📈

Prefrontal and lateral entorhinal neurons co-dependently learn item–outcome rules - Nature

The bidirectional loop circuit between layers 5/6 of the lateral entorhinal cortex and the medial prefrontal cortex encodes item–outcome associative memory in mice.

First episode of our new podcast is out! Episode 1: "What is Intelligence?" It's hard to fully answer this question, but we had some great discussions about it with superstars Alison Gopnik and John Krakauer. Give it a listen! complexity.simplecast.com/episodes/nat...

Bluesky now has over 10 million users, and I was #4,792,410!

Today's Q for discussion by the #NeuroAIarxiv.org/abs/2301.00234 Do we have clear evidence of rapid learning with no synaptic changes in the brain? If so, when/where?

A Survey on In-context Learning

With the increasing capabilities of large language models (LLMs), in-context learning (ICL) has emerged as a new paradigm for natural language processing (NLP), where LLMs make predictions based...

1/ Here's a critical problem that the #neuroai field is going to have to contend with: Increasingly, it looks like neural networks converge on the same representational structures - regardless of their specific losses and architectures - as long as they're big and trained on real world data. 🧠📈 🧪