biblewheel.com

Kyle wrote: Sun Apr 26, 2026 1:51 pm Very interesting that I too thought to do this exact thing. I had a feeling there could be relations to each section of the brain and the canonical divisions. There are some very noteworthy scores there! Can you share the exact prompt you gave Grok to generate this result? I want to try prompting Claude Opus with it as well to see what it concludes. If the results are similar then there may be significance to this.

It was a series of prompts. It probably helps to follow them to set up the context. Here they are:

PROMPT 1:
What do you think of this page? Please try looking at the linear and circular representations. And change the min refs per arc to see what happens.
 
https://biblewheel.com/resources/code/t ... graph.html


PROMPT 2:
Do you see how the seven canonical divisions have more references within themselves relative to connections to other regions? Compare this with the connectome of the human brain (attached) from this page: https://brainum.mox.polimi.it/

I attached this image from that article:
PROMPT 3:
Create a 7 x 7 matrix to compare all possible pairs of the seven canonical divisions & brain lobes. Summarize the main theme/function of each division/lobe and then compare each pair and give a score in the corresponding cell that measures how tightly they cohere. I want to know if there is a strong correlation that links each division to its own brain lobe. Use this page to get more info on the seven divisions: https://www.biblewheel.com/Canon/SevenfoldCanon.php

PROMPT 4:
Yes, please reorder the matrix to show the best diagonal. Ensure that the mapping is one-to-one (no two divisions mapped to the same lobe).

That's it! It will be interesting to see what you find.

I fed the same prompts into ChatGPT and got similar results, but not exactly the same.
I then had both create difference matrices so they could compare and explain their differences. Here is what Grok produced:

Just ran your prompts on Claude Opus 4.7. Similar but different results in column ordering. If you compare manually with your results comparing each division to it's lobes value between matrices the numbers are very similar for each pair but the diagonal differs. You can read the whole chat here: https://claude.ai/share/e5eda55c-3743-4 ... c4a93c877c

Kyle wrote: Sun Apr 26, 2026 3:10 pm Just ran your prompts on Claude Opus 4.7. Similar but different results in column ordering. If you compare manually with your results comparing each division to it's lobes value between matrices the numbers are very similar for each pair but the diagonal differs. You can read the whole chat here: https://claude.ai/share/e5eda55c-3743-4 ... c4a93c877c

Very interesting! You could just copy/paste the matrices from Grok and ChatGPT and have Claude compare them.

And here's the results from Google's Gemini 3.1 Pro. This seems to be an outlier in all the AI models seeing that there is no diagonal score lower than 9 lol.