HN Theater @HNTheaterMonth

You might find https://www.youtube.com/watch?v=DpXy041BIlA interesting. He creates a tournament of "imperfect" engines.

Tom's really mastered his skill of casual misleading, he adds these "wait, what?" moments to every video. A more notorious example will be his explanation of how optical reflection works, check out his "30 Weird Chess Algorithms" (https://youtu.be/DpXy041BIlA) around 10:20.

Edit: spelling.

Tom7 is brilliant for anyone that hasn't seen him before. His Reverse Emulation video (https://youtu.be/ar9WRwCiSr0) and Weird Chess Algorithms (https://youtu.be/DpXy041BIlA) are a must watch. His dead pan delivery coupled with the ridiculous amount of effort he puts into things that don't _really_ matter is honestly inspiring.

⬐ davidfactorial

I think the Reverse Emulation was one of the most technically impressive hacks I've ever seen.

⬐ fps_doug

I'm anticipating his video every April 1st, and was already fearing he might have skipped this year when it wasn't uploaded in time. One of the people with the perfect mix of funny, smart and bat-shit crazy.

⬐ rrauenza

We first enjoyed his AI based video game players... they're also really good!

⬐ jrmann100

Stumbled upon his work on HN with "Uppestcase and Lowestcase Letters":

https://news.ycombinator.com/item?id=26667852

tom7 / suckerpinch is great, I love his "30 Weird Chess Algorithms" video:

https://www.youtube.com/watch?v=DpXy041BIlA

It's unknown what role if any BPE plays. I was surprised to discover that the final probability of a move is equal to the probability of each token from the root prompt, i.e. even though "1.Nf3 e5" is encoded as ['1', '.', 'N', 'f', '3', ' e', '5'] the probability of e5 seems unaffected by the fact that Nf3 is 3 tokens as opposed to one.

You're right that coming up with a token mapping could help things. It's a bit tricky to do that right now. Your options for fitting a custom vocab seems to be "use sentencepiece to fit a vocab, then modify the gpt-2 codebase to use the sentencepiece library for decoding".

I am honestly not sure if the output of sentencepiece is compatible with traditional encoders. What I mean is, it doesn't seem to generate an encoder.json + vocab.bpe file. It seemed to be some other kind of format. So I'm not sure if the tooling that has evolved around OpenAI's encoder format would be applicable there. I really don't know, though.

According to this slatestarcodex comment, someone got superior results on solely algebraic notation (which looks like g1f3 instead of Nf3): https://www.reddit.com/r/slatestarcodex/comments/el87vo/a_ve...

Another extension that might help is to periodically inject the full FEN board state. This was the format we were going to try next, which injects the full FEN after every move: https://gist.github.com/shawwn/318606c112774ad070f94de9c8288...

I'm so happy to get to work with GPT-2 1.5B. It's been a lot of fun to train.

By the way, if you like this kind of thing, you'll love Elo World. https://www.youtube.com/watch?v=DpXy041BIlA

Only somewhat related, but if you found these ideas amusing, you'd probably enjoy watching tom7's video on unusual chess engine strategies: https://youtu.be/DpXy041BIlA

tom7 videos are pure gold. I particularly liked his chess video: https://www.youtube.com/watch?v=DpXy041BIlA

I had just looked up this wiki article a couple days ago based on this video:

https://youtu.be/DpXy041BIlA, "30 Weird Chess Algorithms: Elo World"

An intricate and lengthy account of several different computer chess topics from my SIGBOVIK 2019 papers. We conduct a tournament of fools with a pile of different weird chess algorithms, ostensibly to quantify how well my other weird program to play color- and piece-blind chess performs. On the way we "learn" about mirrors, arithmetic encoding, perversions of game tree search, spicy oils, and hats.

Is this also why it showed up just now?

Suckerpinch (Tom) recently had a video[1] about predicting the position of the oponents pieces from knowning just white. I don't think he's aware of this. Great video anyway.

He trained an NN to predict the pieces and then used Stockfish on the resulting position.

[1] https://www.youtube.com/watch?v=DpXy041BIlA

⬐ NoodleIncident

Yeah, the "pawn tries" rule in particular reminded me a lot of his "blind spycheck" algorithm in that video. If you are allowed to rank every move and perform the first legal one, you can get away with some interesting stuff.

Tom7 made a video also: https://youtu.be/DpXy041BIlA

If you haven’t seen them his other videos are well worth watching as well.

⬐ lemoing

I came across his channel recently and have been watching as many videos as I can. His channel reminds me why I got into computer science in the first place.

⬐ pronoiac

He's a mad genius! I like his compiler that generates executables made up only of printable characters: https://www.youtube.com/watch?v=LA_DrBwkiJA / https://news.ycombinator.com/item?id=16312317

He's super playful and awesomely smart!

⬐ aisha123

https://www.youtube.com/watch?v=SBjLw3_47jg