nnue
maia-chess
nnue | maia-chess | |
---|---|---|
3 | 97 | |
44 | 897 | |
- | 2.1% | |
0.0 | 1.6 | |
over 3 years ago | 10 months ago | |
Python | ||
- | GNU General Public License v3.0 only |
Stars - the number of stars that a project has on GitHub. Growth - month over month growth in stars.
Activity is a relative number indicating how actively a project is being developed. Recent commits have higher weight than older ones.
For example, an activity of 9.0 indicates that a project is amongst the top 10% of the most actively developed projects that we are tracking.
nnue
-
AlphaFold 3 predicts the structure and interactions of all of life's molecules
The original NNUE paper cites AlphaZero[0]. The architectures are different because NNUE is optimized for CPUs and uses integer quantization and a much smaller network. I don't think one could credibly claim that it would have come about if not for Google making so much noise about their neural network efforts in Go, Chess and Shogi.
0: https://github.com/asdfjkl/nnue/blob/main/nnue_en.pdf
-
Hey, what ever happened to Alpha Zero?
The NNUE paper (original, English translation) came out in April 2018 and the first NNUE engine in May, so it's only few months after the A0 paper and I think it's too close to be an influence.
- Efficiently updatable neural networks for computer shogi [pdf]
maia-chess
-
Grandmaster-Level Chess Without Search
This was studied with the Maia series of bots. See:
https://github.com/CSSLab/maia-chess
-
Chess-GPT's Internal World Model
There is a very interesting project on this exact problem called Maia, which trains an engine based on millions of human games played on Lichess, specifically targeting varying levels of skill from 1300 to 1900 Elo. I haven't played it myself, by my understanding is that it does a much better job imitating the mistakes of human players. https://maiachess.com
- A chess terminal user interface implementation
- Interested in going to my first tournament and was wondering if there is a place where I can review players classical games around 1700 rating so I can get an idea of what is expected and their strengths?
-
Scientists claim >99 percent identification rate of ChatGPT content
In general the goal of these progrms is to win, not to pretend to be human.
In chess, there is a project called Maia which aims at predicting the human move rather than the best move. Even then it blunders less than humans of a similar rating, so it can still be detected.
https://maiachess.com/
-
Most human engine to play against?
Maia Chess
-
Do you have to buy Maia to use it offline?
It's on GitHub, you just need an interface like Nibbler.
-
Anyone know if there are chess AIs trained like chatGPT, as a move predictor instead of a move maximizer as most have been (I think)?
That's exactly what Maia Chess is designed to do. https://maiachess.com/
-
Can a 2000 elo bot blunder like this?
If you wanna play a more human like bot check out Maia https://maiachess.com/
-
The Q&A Megathread for new and beginner chess players
If you want to play an engine who plays a lot more like a human, you should try Maia Chess.
What are some alternatives?
Stockfish - A free and strong UCI chess engine
Stockfish - UCI chess engine
stockfish - Integrates the Stockfish chess engine with Python
lc0 - The rewritten engine, originally for tensorflow. Now all other backends have been ported here.
Winter - UCI Chess Engine
zahak - A UCI compatible chess AI in Go
lila - ♞ lichess.org: the forever free, adless and open source chess server ♞
MongoDB - The MongoDB Database
Auto-Chess - A chess bot that automatically calculates the best moves and plays them for you
pgnToFen - Takes PGN as input and returns FEN
product-extraction-benchmark