News for hackers. Stuff that matters.
Re:Chess-playing neural network LC0 beats Stockfish in 100-g... by ngcc_hk on Wednesday May 29, 2019 @06:50AM (#20038627)
Some of the discussion seems forget that the lc0 use neural network. There is essential no logic and there is no mtcs in the core part. Just play.

And it is called zero ad it does not read any human game, like alpha go zero.

It is just play and play to train the neural network ... you may say it want to find a function by itself that play the game. And the function is in the weights.

There are no comments on this comment yet.