wholemoley

Joined 8 years ago

Karma
49
Submissions
17
Member since
Jul 22
Lurker, Interested in AI, RL, GA and your recipes for jujubes.
View on Hacker News

Recent submissions

Comment · 7 years ago
As I mentioned in the other post, the curiosity bots can help with this.

They're rewarded by exploring, not by values in game.

Maybe alone they're not enough, but in conjunction with other things, I bet we could beat Zelda. I had the bot exploring enough to find the first dungeon of LoZ.

Comment · 7 years ago
This person https://github.com/pathak22/noreward-rl and his curiosity AI https://github.com/openai/large-scale-curiosity can do some cool stuff.

I used it to train an AI to play Mario Kart ( https://www.youtube.com/watch?v=A8oSnh0M864 )

I also had it playing The Legend of Zelda (got as far as finding the first dungeon, but with more power, I'm certain it could explore the whole map.).

Comment · 7 years ago
Comment · 8 years ago
Mark's Handbook for Mechanical Engineers, and AREMA.
Comment · 8 years ago
Yeah, topology and weights. It's highly subject to initial conditions. You almost need another NEAT network to evolve the initial conditions. I believe it's turtles all the way down.
Comment · 8 years ago
Comment · 8 years ago
Comment · 8 years ago
Comment · 8 years ago
Comment · 8 years ago
Comment · 8 years ago
Ok, I'll clean it up and post it.
Comment · 8 years ago
Alrighty, I'll clean it up and post it.

Live from the official Hacker News API

A modern reader for Hacker News · SvelteKit + Skeleton