Stories by georgehill

Meta FAIR: Memory Mosaics

Do Llamas Work in English? On the Latent Language of Multilingual Transformers

It Only Takes Two Weeks [video]

Fei-Fei Li: Stanford's Natural Language computing lab has only 64 GPUs

Mikhail Parakhin No Longer Working on Copilot at Microsoft

Google CEO Sundar Pichai and the Future of AI [video]

OpenAI: Model Spec

How are embeddings affecting traditional text search?

Avoiding the Frankenstein Prompt

Kolmogorov–Arnold Networks: Alternative to Multilayer Perceptrons.

Scalable Bayesian Inference in the Era of Deep Learning

Kolmogorov-Arnold Networks (Kan)

Kolmogorov-Arnold Representation Theorem

LLM Control Theory Seminar (April 2024) [video]

React 19 Beta

Scaling Laws vs. Model Architectures: How Does Inductive Bias Influence Scaling?

The "it" in AI models is the dataset

Snowflake Arctic: LLM for Enterprise AI — Efficiently Intelligent, Truly Open

OpenAI: Training LLMs to Prioritize Privileged Instructions

1
2
3
4
5
6
7
8
9
10