61–70 of 71 posts
Earlier quoted context omitted.
The clever titles are more brandable. See the citations for "Attention is all you need" or "Chinchilla limit" versus more mundane titles.
The Chinchilla paper was called "Training Compute-Optimal Large Language Models", which is exactly on point.
Seems to me like the whole history of neural nets is basically crafting models with well-behaved gradients to make gradient descent work well. That, and models that can achieve high utilization of available hardware. The surprising thing is that models exist where the gradients are so well-behaved that we can learn GPT-4 level stuff.
There's plenty of interesting neural network designs out there but they're being overshadowed by transformers due to their recent success. I personally thing that the main reason transformers work so well is because they actually step away from the multi layer perceptron stuff and introduce some structure and in a way sparsity.
But it's equally important to create architectures that allow efficient backpropagation of errors.
It does seem like transformers are pretty good at both, already.
I kind of hope we're not getting much something radically better anytime soon, because it seems like AGI is already approaching faster than we can prepare for.
Then again, I would expect that someone somewhere is already using transformer based networks to develop some brand new architecture that does in fact provide such a leap.
Global optimization techniques which don't rely on gradients seems theoretically superior in all instances, except that we haven't found super fast ways to run these kinds of optimizers. The cartpoll demo famously tripped up derivative based reinforcement learning for awhile.
Could you point me toward any more info or keywords on "the cartpoll demo famously tripped up derivative based reinforcement learning for awhile"? It sounds like an interesting bit of history but my google searches are only bringing up python tutorials.
The reason for this is that the algorithm doesn't like to have to "spend" energy, reducing its score. Without huge amounts of trickery to get the gradient descent algorithm to stop getting stuck in the center, this is never solved - due to using a local optimizer for a global optimization problem (finding good weights in a NN)
Earlier quoted context omitted.
Could you point me toward any more info or keywords on "the cartpoll demo famously tripped up derivative based reinforcement learning for awhile"? It sounds like an interesting bit of history but my google searches are only bringing up python tutorials.
Chatgpt: What does this refer to: cartpoll demo famously tripped up derivative based reinforcement learning The phrase "cartpole demo famously tripped up derivative-based reinforcement learning" is likely referring to a classic problem in the field of reinforcement learning, which involves balancing a pole on a cart. The pole is attached to the cart via a hinge, and the goal is to keep the pole upright by moving the…
Earlier quoted context omitted.
That's the expansion. What does it represent?
Given the context of the original statement, the Taylor series expansion represents the statistical learning process of any topic anywhere.
Earlier quoted context omitted.
Given the context of the original statement, the Taylor series expansion represents the statistical learning process of any topic anywhere.
I think you mean loss function, not process. And I'm failing to make any connection between intelligence and polynomial approximations to loss functions.