Mathavan discrepancies

[ekiefl]

Hey @tailuge, I was wondering what you think about Mathavan's overspin result in Figure 9:

It's an obvious divergence from your recapitulated diagrams and I was wondering if you knew why.

[tailuge]

Its a good point. At the time I put this down to possible mistake in the paper? I've just tried upping the number of iterations in the numeric solution and it makes no change to my plot. Does your implementation match the paper?

[ekiefl]

My implementation is just a conversion of your typescript code, so it looks just like yours.

It's possibly an error on their end. It's suspicious that the final data point is significantly lower than the second last data point.

[tailuge]

Ok. Good luck with the conversion (use an LLM for it?). I'm sure there are mistakes lurking, do let me know.

On the LLM front I got a multiplayer lobby coded up mostly by LLM

https://scoreboard-tailuge.vercel.app/test.html (3 iframes to test it)

of course it lacks actual users...