4 ms·
That's addressed in the article: > The team spent half a year trying to understand how the neural network had gotten so wise. They checked to make sure the alg
by lucb1e 5y ago
That's addressed in the article:
> The team spent half a year trying to understand how the neural network had gotten so wise. They checked to make sure the algorithm hadn’t just found some way to infer the density from the coding of the simulation rather than the galaxies themselves.
- nieve 5y agoExcept that they acknowledge its performance was dependent on the details of the simulation and failed with a different recipe for making universes, so in fact it was dependent on not just the code, but the settings.