I've been working on something similar, implementing a relational language on top of a tensor library[0].<p>Mathematically, einsum and database joins are the same thing, just over different semirings (real numbers for einsum, booleans for databases). A lot of papers about datalog explore this sort of thing in more depth. In particular, Dyna[1] might be interesting.<p>[0]: <a href="https://arxiv.org/abs/2509.22614" rel="nofollow">https://arxiv.org/abs/2509.22614</a>
[1]: <a href="https://dyna.org/" rel="nofollow">https://dyna.org/</a>
Somewhat more reliable than implementing SQL in neural networks.
I'm just going to go back to making my CRUD endpoints...<p>Jokes aside, sounds really impressive, though I only understood about 10% :D
initially rolled my eyes at "neural networks in SQL," but after reading the code I came away impressed<p>basically it comes down to using relational algebra as the IR, letting a database optimizer reason about tensor programs
Neat! Feels analogous to "X runs Doom" demos (but NN).
Just today I saw this implementation of DOOM in SQLite using a recursive CTE for a simple ray tracer: <a href="https://github.com/petergpt/doomql" rel="nofollow">https://github.com/petergpt/doomql</a>
Totally. I can’t wait to take this to <a href="https://hytradboi.com" rel="nofollow">https://hytradboi.com</a>
Why? lol
[dead]
[dead]