AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
How Does a Neural Operator Look Like?
In the natural image domain, we want to all the time filter out. All the pixels is just too high dimensional and mostly useless information. But that's different in so many of the scientific simulations like fluid dynamics because you cannot just filter out or smooth out the higher resolution information. And so the way we went about overcoming this fundamental drawback is to come up with now framework called neural operator. So once you've trained the model at using some training data at certain resolution, at testing time or at inference time, you can test it at any resolution.