To argue with a person who has renounced the use of reason is like administering medicine to the dead. - Thomas Paine

axios.com/spotify-patent… Weird and scary?

An overdue reminder of what can happen when you go to the Moon.

After completing the Cognitive Science program at McGill University with First Class Honours, I am thrilled to announce that I will be joining the Human and Machine Learning Lab as well as the Computation and Cognition Lab at NYU as a PhD student in the f…lnkd.in/dz-Pg_K8

New blog post out! solimlegris.com/anglais/2022/0…

It is widely claimed that DNNs are the “best” models of human vision. Our article entitled “Deep Problems with Neural Network Models of Human Vision” challenges this view and proposes a way forward. Thread: 1/25 psyarxiv.com/5zf4s/

Much of our lives is shaped by the people we choose to hang out with. And that's why you should read books. Good ones. Easiest way to spend quality time with interesting people of your choice.

I could never get used to the fact that English uses 1-indexing when referring to floors in a building. I much prefer 0-indexing, like in French or in Python.

Today (now!), Solim LeGris explains how we predict insight using behavioural data in a physical reasoning game. Check out poster #CogSci2024 P3-LL-867 at 1pm today! Learn about the project here: exps.gureckislab.org/e/blue-giganti…

One inspiration for ARC-AGI solutions is the psychology how humans solve novel tasks. A new study by todd gureckis Brenden Lake Solim LeGris Wai Keen Vong @ NYU explores human performance on ARC, finding that 98.7% of the public tasks are solvable by at least 1 MTurker.

We have a new preprint led by Solim LeGris and Wai Keen Vong (w/ Brenden Lake ) looking at how people solve the Abstraction and Reasoning benchmark. arxiv.org/abs/2409.01374 ARC is one of the most challenging & long standing AGI benchmarks and the focus of a large $$ prize ARC Prize 🧵

In honor of the late Freddie Mercury's 78th bday, a programmer's ode to Queen: bit.ly/2buA9Mq

do large-scale vision models represent the 3D structure of objects? excited to share our benchmark: multiview object consistency in humans and image models (MOCHI) with Stephanie Fu Yutong Bai Thomas O'Connell @_yonifriedman Nancy Kanwisher @[email protected] Josh Tenenbaum and Alexei Efros 1/👀

Humans still outperform AI on visual reasoning tasks, according to a new study by Solim LeGris, Wai Keen Vong, Brenden Lake, and todd gureckis. Despite advances, the top AI models are still significantly worse than humans on the ARC Prize's ARC benchmark. nyudatascience.medium.com/human-intellig…

Can neuro-inspired ANN architectures be useful for motor control in quadruped robots? We translate neural circuits in the limbs and spinal cord of mammals into an ANN architecture controlling quadruped locomotion. w/ Venkatesh, Lerrel Pinto, @neurograce

The YouTube video I want to watch is any highly rated, 1hr long, information dense lecture on anything esoteric and the algorithm just doesn’t get it. It’s too content-driven and too narrow-minded

There are infinitely many ways to write a program. In our new work, we show that training autoregressive LMs to synthesize programs with sequences of edits improves the trade-off between zero-shot generation quality and inference-time compute. (1/8) w/ Rob Fergus Lerrel Pinto

Classic challenges for neural nets focus on human strengths that are model weaknesses. With Kazuki Irie, we discuss 4 cases of neural nets overcoming weaknesses with practice (meta-learning). A framework for bringing machine and human intelligence closer? arxiv.org/abs/2410.10596

Of those who made the critical contributions on the path from the first programmable computers to today's LLMs ~ half are psychologists: Rosenblatt (perceptron) Rumelhart, McClelland, Hinton (backprop, deeper nets) Elman (RNN, autoregressive training)

.Solim LeGris and Wai Keen Vong estimated that average human performance on ARC is about 64% correct (on the public eval set). This would make o3 clearly better than the average human. Notably, almost all tasks were solvable by at least one person. arxiv.org/abs/2409.01374