This talks about representing numbers as graphs whilst showing that the graphs can appear to be reminiscent of leaves, not that the numbers are leaf nodes of graphs, which is where I thought it was going at first.

Is there a paper available?

Tough question (for me). Assuming the model is producing its own queries, am I wrong to wonder how it's fundamentally different from human reasoning?

It could just be programmed to follow up by querying itself with a prompt like "Come up with arguments that refute what you just wrote; if they seem compelling, try a different line of reasoning, otherwise continue with what you were doing." Different such self-administered prompts along the way could guide it through what seems like reasoning, but would really be just a facsimile thereof.

Maybe the model doesn't do multiple queries but just one long query guided by thought tokens.

People are saying it's about rhyming but I don't think so:

0, done, 2, free, 4, hive, 6, jeven, 8, line, 10, neleven, 12, pirteen, ...

>The reward signal is analogous to pleasure and pain for biological systems...

Anyone knowing the basics of Reinforcement Learning will know that this is misleading and incorrect.

One reason spices were treasured may be because back then, European food was much more bland that it is today. Think of all the foods of American origin which weren't known about (potatoes, sweet potatoes, corn, chili peppers, cocoa etc).

Barely Adequate Rail Transit

I hope this is an academic exercise because if absolute zero entails no air travel, no shipping, no lamb or beef, amongst other prohibitions, it also entails violence.