unknown

physical systems (orbits/fluid mechanics) may look complex, but are often governed by simple equations/few parameters. can current self-supervised methods learn the underlying physics?

our new paper finds that learning in latent space may be the key!

🧵

<p>physical systems (orbits/fluid mechanics) may look complex, but are often governed by simple equations/few parameters. can current self-supervised methods learn the underlying physics?

our new paper finds that learning in latent space may be the key!

🧵</p>

Yann LeCun

A retweet referencing the idea that physical systems like orbits and fluid mechanics, despite appearing complex, are governed by simple equations and few parameters, with a suggestion that this principle may apply to other domains.