News for hackers. Stuff that matters.
Re:Scallop – A Language for Neurosymbolic Programming by JFingleton on Saturday March 22, 2025 @04:00AM (#43444281)
A bit over my head - but can't Prolog achieve similar results?
Re:Scallop – A Language for Neurosymbolic Programming by sgt101 on Saturday March 22, 2025 @04:15AM (#43444323)
Anything can do anything else given enough time and power, but I think: no, not without shenanigans. This has primitives for interfacing to nn's including foundation models so you can ask it (for example) to label images of cats and dogs using clip, then you can reason over the results.

So it's intended to combine nn reasoning and logical reasoning cleanly.

Re:Scallop – A Language for Neurosymbolic Programming by nurettin on Saturday March 22, 2025 @12:33PM (#43446901)
Scallop's examples have syntax for probabilistic programming, so probably not.