AlphaGo Zero: Learning from scratch

Re:AlphaGo Zero: Learning from scratch by panic on Wednesday October 18, 2017 @02:21PM (#15501556)

Is AlphaGo Zero the first Go program without special code to read ladders? I'm curious how a pure neural net can read them, given how non-local they are.

[ Parent | on HN ]

Re:AlphaGo Zero: Learning from scratch by thefalcon on Wednesday October 18, 2017 @05:16PM (#15503198)

The concept of locality is nothing but a human weakness in Go, the best AI must read the whole board with every move.

EDIT: From the paper: "Surprisingly, shicho (“ladder” capture sequences that may span the whole board) – one of the first elements of Go knowledge learned by humans – were only understood by AlphaGo Zero much later in training" I'm surprised by the author's use of the word "Surprisingly" here.

[ Parent | on HN ]

Re:AlphaGo Zero: Learning from scratch by waqf on Wednesday October 18, 2017 @07:07PM (#15504062)

AlphaGo is still based around layers of 3×3 local convolutions.

That represents a strong assumption about locality in the network design. I would expect AlphaGo to perform poorly on the game "Go with the vertices randomly permuted".

[ Parent | on HN ]

Re:AlphaGo Zero: Learning from scratch by Palptine on Thursday October 19, 2017 @10:19AM (#15508088)

Well they didn't use inception. If their inception units have, say, a 7x7 conv, then ladders will probably be found much earlier.

[ Parent | on HN ]

Re:AlphaGo Zero: Learning from scratch by gizmo686 on Wednesday October 18, 2017 @05:25PM (#15503289)

>Surprisingly, shicho (‘ladder’ capture sequences that may span the whole board)—one of the first elements of Go knowledge learned by humans—were only understood by AlphaGo Zero much later in training. [0]

[0] https://www.nature.com/nature/journal/v550/n7676/full/nature...

[ Parent | on HN ]