Today we put a message at the top of Instagram in the US making it clear that votes are still being counted. 🗳
This is a moment of uncertainty for a lot of people, and we won’t have clarity for a few days. In the meantime it’s critical that people have accurate information.
We have studied recipes for large-scale open domain chatbots, and are releasing 90M, 2.7B and 9.4B parameter models with SOTA results.
Paper: https://t.co/ORY7zemRdE
Project: https://t.co/L2lcjO5Cj7
Blog post: https://t.co/uMSAJMvLii
Emissions de gasos d'efecte hivernacle en els aliments: el que importa és "què" menges. L'origen (és a dir, el transport) és gairebé irrellevant. Via @OurWorldInData
1. I worked with Anna Zaitsev (Berkely postdoc) to study YouTube recommendation radicalization. We painstakingly collected and grouped channels (768) and recommendations (23M) and found that the algo has a deradicalizing influence.
Pre-print:
https://t.co/1NneHDnKHD
🧵
Hem parlat amb en @ros_xavi de matemàtiques, de les equacions que mouen el món, de la tímida política de recerca espanyola... I del llaç groc. Tot sentit comú.
Incredible work on Bayesian Optimization by my colleagues on Facebook's Adaptive Experimentation team. This is the next generation of A/B testing. https://t.co/v1NLsCUJE8