Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Grandmaster-level chess without search

github.com

Post reply on HN
Print view View original

81–90 of 171 posts

Re: Grandmaster-level chess without search

#81

Post by Davidzheng »

famouswaffles wrote: post #39
imranhou wrote: post #35

From the page: "We also show that our model outperforms AlphaZero's policy and value networks (without MCTS) and GPT-3.5-turbo-instruct." Why compare this to GPT-3.5-turbo-instruct? Is that near SOTA in this space?

As far as anyone knows, 3.5-turbo-instruct is the best chess playing (certainly it was at the time of the paper) LLM. About 1800 Elo and < 0.1% Illegal move rate. It's unclear why it was so much better than 4 (lack of RLHF?, Data?) and I don't know if anyone has bothered to test 4o similarly but it was pretty big news online at the time.

OA definitely purposefully trained its chess strength

Re: Grandmaster-level chess without search

#82

Post by Davidzheng »

janalsncm wrote: post #11
alfalfasprout wrote: post #5

Earlier quoted context omitted.

The thing is classical chess (unlike eg; go) is essentially "solved" when run on computers capable of extreme depth. Modern chess engines play essentially flawlessly.

Chess is not “solved”. Solved doesn’t mean computers can beat humans, it means for any chess board position we can tell whether white wins, black wins, or the game is drawn with perfect play. We would know if the starting position was drawn, for example. No computers now or in the foreseeable future will be capable of solving chess. It has an average branching factor over 30 and games can be over 100 moves.

There's strong solved and weak latter only needs to be unbeatable from starting. Now it's definitely not probable that SF isn't beatable from initial but honestly it's not impossible. The drawing margin is pretty big for engines

Re: Grandmaster-level chess without search

#83

Post by Davidzheng »

solveit wrote: post #7
alfalfasprout wrote: post #5

Earlier quoted context omitted.

The thing is classical chess (unlike eg; go) is essentially "solved" when run on computers capable of extreme depth. Modern chess engines play essentially flawlessly.

We really have no way to know this. But I would be very surprised if modern chess engines didn't regularly blunder into losing (from the perspective of a hypothetical 32-piece tablebase) positions, and very very surprised if modern chess engines perfectly converted tablebase-winning positions.

Ofc they do but the more interesting question for weak solved is whether they do in mainline positions (like mainline Berlin, mainline petroff, etc) where you can hold equality in many ways and engines are printing 0.0 everywhere

Re: Grandmaster-level chess without search

#84

Post by Davidzheng »

janalsncm wrote: post #18
solveit wrote: post #7

Earlier quoted context omitted.

We really have no way to know this. But I would be very surprised if modern chess engines didn't regularly blunder into losing (from the perspective of a hypothetical 32-piece tablebase) positions, and very very surprised if modern chess engines perfectly converted tablebase-winning positions.

The fact that TCEC games aren’t all draws suggests that computers aren’t perfect. Stockfish loses to Leela sometimes for example.

Most TCEC starting positions are borderline lost

Re: Grandmaster-level chess without search

#85

Post by Davidzheng »

KK7NIL wrote: post #32
solveit wrote: post #7

Earlier quoted context omitted.

We really have no way to know this. But I would be very surprised if modern chess engines didn't regularly blunder into losing (from the perspective of a hypothetical 32-piece tablebase) positions, and very very surprised if modern chess engines perfectly converted tablebase-winning positions.

We do know this, there are many positions (primarily sharp middle game one's) where SF/lc0 will significantly change their evaluation as they go deeper. This problem gets better the more time they spend on one position but it's an inevitable consequence of the horizon effect and it's why (except for 8 pieces or less), chess is far from solved.

Far from strongly solved but i would wager current SF will not lose half of its white games against any future engine

Re: Grandmaster-level chess without search

#86

Post by Davidzheng »

primitivesuave wrote: post #13
alfalfasprout wrote: post #5

Earlier quoted context omitted.

The thing is classical chess (unlike eg; go) is essentially "solved" when run on computers capable of extreme depth. Modern chess engines play essentially flawlessly.

This is accurate for endgames only. In complicated positions, there is still room for improvement - the recent game of lc0 vs stockfish where lc0 forced a draw against an impending checkmate is a good example. There is currently no way for a chess engine searching a massive game tree can see how an innocuous pawn move enables a forced stalemate 40 moves down the line.

Honestly SF plays better in middle game positions on average I would guess. I think usually there's a bigger draw margin in middle games

Re: Grandmaster-level chess without search

#88

Post by hilux »

pertymcpert wrote: post #74
bicsi wrote: post #70

They built a dumber clone of Stockfish, and they call it ‘zero’ for some reason. What is the meaning behind ‘zero’ anyways? It used to refer to zero-shot, but now it seems like it’s just a marketing term.

Zero doesn't mean zero shot learning. It was coined by Deepmind for AlphaGo Zero where they used zero human input into the training data. It was trained entirely by playing against itself.

And they gave Sir Demis a Nobel Prize!

Re: Grandmaster-level chess without search

#89

Post by pgt »

salamo wrote: post #43
tzs wrote: post #20

OT: what's the state of the art in non-GM level computer chess? Say I want to play chess with an opponent that is at about the same skill level as me, or perhaps I want to play with an opponent about 100 rating points above me for training. Most engines let you dumb them down by cutting search depth, but that usually doesn't work well. Sure, you end up beating them about half the time if you cut the search down enoug…

I built something like this. It works as long as you're not too high-rated: chessmate.ai. Once players get higher rated it is more difficult to predict their moves because you need to model their search process, not just their intuitive move choice. It's also possible to train on one player's games only so that it is more personalized. It uses a similar approach to Maia but with a different neural network, so it had…

Hello! I built a Chess AI also named Chessmate in high school (2005) that made it into a Minecraft mod 10 years later: http://petrustheron.com/posts/chessmate.html

Java source code here: https://github.com/theronic/chessmate

Re: Grandmaster-level chess without search

#90

Post by amunozo »

chvid wrote: post #77

But the gigantic synthetic dataset that is used for training is created with plenty of traditional search. So it is all a bit silly but I guess cool none the less ...

It's a knowledge distillation. You can then use this smaller, more efficient models instead of the larger one.
Post reply on HN