I should say that the project had been initiated earlier by @EXGRV who wrote the first version of the code.
@ylecun
-
Meta FAIR LLM Projects: OPT, Galactica, and Zetta Development
By
–
You misread.
There had been multiple LLM projects within FAIR for years. Some were open sourced as research prototypes (e.g. OPT175B, Galactica, BlenderBot…).
In mid-2022, FAIR started a large LLM project called Zetta, which was still going in late 2022 when ChatGPT came out.
A -
French pioneers: Model predictive control and system identification history
By
–
Hey dudebot, you wanna play that game?
The first publication about model predictive control with system identification (learned world model) is from a bunch of French dudes in 1978. You and I were still in high school.
They proposed a method called IDCOM, which I learned -
AGI emergence will be gradual not sudden
By
–
– the emergence of "AGI" (whatever your definition) will not be an event. It will be progressive.
– once it appears somewhere, it will be reproduced by many in a relatively short time.
– you say that AI innovations originated in the US. It's not entirely true. Although many -
DINO-World Model: Planning with JEPA Framework
By
–
New work: DINO-World Model.
— Yann LeCun (@ylecun) 31 janvier 2025
Planning with JEPA.
⬇️⬇️⬇️ https://t.co/ruknaI4LbuNew work: DINO-World Model.
Planning with JEPA. -
Silicon Valley’s Misplaced Superiority Complex Problem
By
–
A common disease in some Silicon Valley circles: a misplaced superiority complex.
-
Crowdsourced Distributed Fine-Tuning: A Year-Long Advocacy
By
–
I've been arguing for something like this for over a year: crowdsourced distributed fine-tuning.
-
Groq’s Llama-3.3 Distilled from DeepSeek-R1 Open Source
By
–
Groq's Llama-3.3 distilled from DeepSeek-R1 would belong to the first line.
#OpenSourceMagic -
Graph Neural Networks: Convolutions as Diagonal Operators in Fourier Space
By
–
The root of graph neural convolutional nets, since convolutions are a Diagonal operator in generalized Fourier space.
-
DeepSeek Market Reactions Unjustified: Analysis
By
–
The market reactions to DeepSeek are woefully unjustified.
Here is why