> Let's follow one example: Nigeria is the most populous country in Africa. In Abstract Wikipedia, this might be stored as: Z27243(Q1033, Q138758272, Q6256, Q15, Z27243K5)
Haha that's like John Wilkins' "Real Character, and a Philosophical Language"
It's not that different from how LLM tokens work, only in a tree structure as opposed to a plain sequence. Having a tree structure makes it easier to formally define rewrite rules (which is key for interpretability), as opposed to learning them from data as LLM do.
For context, this was proposed way back in 2013 (https://meta.wikimedia.org/wiki/Abstract_Wikipedia), when machine translation is just plain bad (and LLMs are only known in academic circles). Surprised that AWiki is now active though.
> Let's follow one example: Nigeria is the most populous country in Africa. In Abstract Wikipedia, this might be stored as: Z27243(Q1033, Q138758272, Q6256, Q15, Z27243K5)
Haha that's like John Wilkins' "Real Character, and a Philosophical Language"
https://en.wikipedia.org/wiki/La_Ricerca_della_Lingua_Perfet... is a great intro to the weird and wonderful world of abstract/universal/ideal/a priori languages.
It's not that different from how LLM tokens work, only in a tree structure as opposed to a plain sequence. Having a tree structure makes it easier to formally define rewrite rules (which is key for interpretability), as opposed to learning them from data as LLM do.
One issue with projects like this just show me what it is on the front page.
Even the featured article section is empty!
Correct, what is this and why is this?
Are they trying to reinvent Cyc?
https://en.wikipedia.org/wiki/Cyc
For context, this was proposed way back in 2013 (https://meta.wikimedia.org/wiki/Abstract_Wikipedia), when machine translation is just plain bad (and LLMs are only known in academic circles). Surprised that AWiki is now active though.
So rather than machine translation... really primitive machine translation with extra steps?
Would you rather use a compiler, or have an LLM generate assembly code based on source code?
The purpose is to establish a new high-level lanugage