2 ms·

It sometimes adds too many pieces: https://noctie.ai/play?newGameDialogState_customGame=true&newGameDialogState_overrideFen=3r2k1%252F1p1b1ppp%252Fp1ppnb2%252F

It sometimes adds too many pieces: https://noctie.ai/play?newGameDialogState_customGame=true&newGameDialogState_overrideFen=3r2k1%252F1p1b1ppp%252Fp1ppnb2%252FP3p2n%252FP1B1PPPq%252F2NpB1QP%252F1P2N3%252FR2R2K1%2520w%2520-%2520-%25200%25201 https://noctie.ai/play?newGameDialogState_customGame=true&ne... I'm pretty sure black should have at most eight pawns.

klysm 2y ago
This feels like a perfect example of just whacking a neural net on something without actually applying and understanding of the domain and validating things.
spacebanana7 2y ago
I wonder whether that could be fixed by using an approach like this [1] [2], to guarantee the model output meets a schema. [1] https://news.ycombinator.com/item?id=37125118 https://news.ycombinator.com/item?id=37125118 [2] https://blog.normalcomputing.ai/posts/2023-07-27-regex-guided-generation/regex-guided-generation.html https://blog.normalcomputing.ai/posts/2023-07-27-regex-guide...
progval 2y ago
Or two bishops of the same color
- cenanozen 2y ago
  That is possible with pawn promotion
GistNoesis 2y ago
There is something wrong with the rendering of the color of pieces : In https://chessdream.ai/?dreamId=ab86d6a2-470a-4238-9181-8b4d3013103b https://chessdream.ai/?dreamId=ab86d6a2-470a-4238-9181-8b4d3... there are two white kings, and queens. The position is correct though when viewed on Noctie. Edit : In fact, the kings and queen are neither black or white but a third color : grey Edit 2: There is probably some issue with the adjust piece count, when opening the same dreamId in multiple windows. As following the same link doesn't have the same number of pieces.
- MereGurudev 2y ago
  Hey, you're right, adjust piece count is a local modification, so opening the same dreamId in multiple windows would let you view the position with different piece counts (and the link to Noctie uses the local state)
MereGurudev 2y ago
Thanks for the feedback! It's a simple modification to guarantee things like this – for this prototype I thought it interesting though to expose the "raw" output of the neural network, including crazy things like sometimes placing two kings of the same color. I even force the neural network to place more pieces than it intends, if you put the piece slider on max, to see what happens (and often the result is very bad)

klysm 2y ago

This feels like a perfect example of just whacking a neural net on something without actually applying and understanding of the domain and validating things.

spacebanana7 2y ago

I wonder whether that could be fixed by using an approach like this [1] [2], to guarantee the model output meets a schema. [1] https://news.ycombinator.com/item?id=37125118 https://news.ycombinator.com/item?id=37125118 [2] https://blog.normalcomputing.ai/posts/2023-07-27-regex-guided-generation/regex-guided-generation.html https://blog.normalcomputing.ai/posts/2023-07-27-regex-guide...

progval 2y ago

Or two bishops of the same color

cenanozen 2y ago
That is possible with pawn promotion

GistNoesis 2y ago

There is something wrong with the rendering of the color of pieces : In https://chessdream.ai/?dreamId=ab86d6a2-470a-4238-9181-8b4d3013103b https://chessdream.ai/?dreamId=ab86d6a2-470a-4238-9181-8b4d3... there are two white kings, and queens. The position is correct though when viewed on Noctie. Edit : In fact, the kings and queen are neither black or white but a third color : grey Edit 2: There is probably some issue with the adjust piece count, when opening the same dreamId in multiple windows. As following the same link doesn't have the same number of pieces.

MereGurudev 2y ago
Hey, you're right, adjust piece count is a local modification, so opening the same dreamId in multiple windows would let you view the position with different piece counts (and the link to Noctie uses the local state)

MereGurudev 2y ago

Thanks for the feedback! It's a simple modification to guarantee things like this – for this prototype I thought it interesting though to expose the "raw" output of the neural network, including crazy things like sometimes placing two kings of the same color. I even force the neural network to place more pieces than it intends, if you put the piece slider on max, to see what happens (and often the result is very bad)