https://github.com/nescitus/LizardBrain
This engine contains the dumbest neural network for chess evaluation that exists. 784x16x1 (the same size as Tomitank, which was a direct inspiration), with extremely dumb tuner caring only for the input layer. Because of working on a single laptop, I took a shortcut and initialized some weights in a way mimicking piece/square tables.
Current iteration earned a dubious distinction of playing at 2000-2100 level while using a 2800 Elo-capable search.
If it proves capable of learning something more, I'll consider resurrecting Rodent with Tomitank-style neural network adjustement of its eval. If not, view it as an educational toy showing what a real neural network does.
The dumbest neural network there is
Moderator: Ras
-
- Posts: 905
- Joined: Mon Jan 15, 2007 11:23 am
- Location: Warsza
-
- Posts: 4822
- Joined: Fri Apr 21, 2006 4:19 pm
- Location: IAȘI - the historical capital of MOLDOVA
- Full name: Silvian Rucsandescu
Re: The dumbest neural network there is
Alf 1.09 64-bit has 2061 Elo (CCRL Blitz). The match is underway.PK wrote: ↑Fri Apr 26, 2024 4:01 pm https://github.com/nescitus/LizardBrain
This engine contains the dumbest neural network for chess evaluation that exists. 784x16x1 (the same size as Tomitank, which was a direct inspiration), with extremely dumb tuner caring only for the input layer. Because of working on a single laptop, I took a shortcut and initialized some weights in a way mimicking piece/square tables.
Current iteration earned a dubious distinction of playing at 2000-2100 level while using a 2800 Elo-capable search.
If it proves capable of learning something more, I'll consider resurrecting Rodent with Tomitank-style neural network adjustement of its eval. If not, view it as an educational toy showing what a real neural network does.

-
- Posts: 12378
- Joined: Thu Mar 09, 2006 12:57 am
- Location: Birmingham UK
- Full name: Graham Laight
Re: The dumbest neural network there is
How about...
1. Take the smartest neural network there is
2. Multiply all its weights by -1
Maybe it could create a new opening: walk the king straight to the opposite side of the board!
1. Take the smartest neural network there is
2. Multiply all its weights by -1
Maybe it could create a new opening: walk the king straight to the opposite side of the board!

Human chess is partly about tactics and strategy, but mostly about memory
-
- Posts: 905
- Joined: Mon Jan 15, 2007 11:23 am
- Location: Warsza
Re: The dumbest neural network there is
New network uploaded, ~2450 Elo
Pawel Koziol
http://www.pkoziol.cal24.pl/rodent/rodent.htm
http://www.pkoziol.cal24.pl/rodent/rodent.htm