About
About this site
A reference work on how AI technologies fit together, written so that every claim can be checked against its source.
This site is a graph: technologies are nodes, and the relationships between them are edges. Each edge records what the relationship is, how confident the record is, when it was last verified, and where the claim came from. A page is only useful if you can follow it back to a primary source, so that is the form the data takes rather than something added afterwards.
It is built for people making a decision - which vector store, which serving stack, whether to fine-tune or retrieve - and who would rather see the differences laid out with citations than read a summary somebody wrote from memory.
Who runs it
The site is maintained anonymously. There is no named author and no claimed credential, because inventing either would be a lie told to a ranking algorithm. What is offered instead is that nothing here asks to be taken on trust: every relationship links to the source it came from, and the method is published on this site rather than described in the abstract.
The site is not affiliated with any of the technologies it catalogues. Nothing is sponsored and no listing can be bought.