Re:Scallop – A Language for Neurosymbolic Programming
by FloorEgg on Saturday March 22, 2025 @01:23AM
(#43443769)
Ever since I learned about category theory and its relationship with symbolic reasoning I've suspected that AGI will come from elegantly combining symbolic reasoning and probabilistic reasoning. This is the first project I've seen that seems to be positioned that way. Very cool.
[ Parent
| on HN ]
Re:Scallop – A Language for Neurosymbolic Programming
by dr_dshiv on Saturday March 22, 2025 @01:33AM
(#43443805)
When LLMs code in order to reason, isn’t that a combination of probabilistic reasoning and symbolic reasoning?
Re:Scallop – A Language for Neurosymbolic Programming
by sigmoid10 on Saturday March 22, 2025 @04:13AM
(#43444319)
Neural networks are actually somewhere in between. They don't directly operate on symbolic expressions or explicit logical rules. And while they rely on probabilistic aspects for training (and sometimes for inference), they rely more on continuous-valued transformations in extremely high dimensional spaces. But if your goal is human-like intelligence, they are a pretty good bet, because we know the human brain also doesn't perform symbolic reasoning at its core and these things only emerge as high-level behaviour from a sufficiently complex system. But it also makes neural networks (and us too) prone to failure modes that you would not see in strictly symbolic reasoning processes.
Re:Scallop – A Language for Neurosymbolic Programming
by mountainriver on Saturday March 22, 2025 @10:21AM
(#43445784)
Yes, this seems to be what the symbolists always forget. We don’t use symbolism like this, we just have very dense neural connections that emerge from scale and approximate it
Re:Scallop – A Language for Neurosymbolic Programming
by suddenlybananas on Saturday March 22, 2025 @10:52AM
(#43446060)
That's not really what symbolists actually argue. Read Fodor and Pylylshyn 1988 to better understand how symbolists view the relationship between symbolic representation and connectionist models. What you're saying is akin to saying there's no point in doing neuroscience or trying to understand neural-networks as its just particle physics deep down.
Re:Scallop – A Language for Neurosymbolic Programming
by sitkack on Sunday March 23, 2025 @05:12AM
(#43451731)
Connectionism and cognitive architecture: A critical analysis
https://www.semanticscholar.org/paper/Connectionism-and-cogn...
Fodor and Pylyshyn's Critique of Connectionism and the Brain as Basis of the Mind https://arxiv.org/abs/2307.14736
Re:Scallop – A Language for Neurosymbolic Programming
by WhitneyLand on Saturday March 22, 2025 @11:09AM
(#43446244)
So if inventing the airplane how long should one stick with a flapping bird wing design?
Re:Scallop – A Language for Neurosymbolic Programming
by nextos on Saturday March 22, 2025 @01:17PM
(#43447200)
That is not a good analogy. Symbolism has given us lots of useful things, including SAT/SMT and theorem provers.
Re:Scallop – A Language for Neurosymbolic Programming
by WhitneyLand on Saturday March 22, 2025 @03:07PM
(#43447842)
The point is that while biology has been a great source of inspiration toward technical advances it’s only a guide, it can’t guarantee there isn’t a better way.
Re:Scallop – A Language for Neurosymbolic Programming
by nextos on Saturday March 22, 2025 @01:15PM
(#43447191)
http://forestdb.org is quite old and includes some toy examples that IMHO elegantly combine symbolic and probabilistic reasoning.
Re:Scallop – A Language for Neurosymbolic Programming
by adastra22 on Saturday March 22, 2025 @10:34PM
(#43450488)
There is a long history of efforts to combine symbolic and connectionist approaches. This is hardly the first!
Re:Scallop – A Language for Neurosymbolic Programming
by mentalgear on Sunday March 23, 2025 @07:43AM
(#43452309)
If you know of other modern NS languages/integrations, please share !
Re:Scallop – A Language for Neurosymbolic Programming
by mentalgear on Sunday March 23, 2025 @07:47AM
(#43452325)
Related: Graph&Category-Theory-based Neuro-Symbolic AI
> The work uses graphs developed using methods inspired by category theory as a central mechanism to teach the model to understand symbolic relationships in science.
https://news.mit.edu/2024/graph-based-ai-model-maps-future-i...
Re:Scallop – A Language for Neurosymbolic Programming
by codethief on Sunday March 23, 2025 @07:49AM
(#43452330)
I know a tiny bit about category theory but nothing about symbolic reasoning. Would anyone mind ELI5ing the connection between the two?
Re:Scallop – A Language for Neurosymbolic Programming
by FloorEgg on Monday March 24, 2025 @09:41PM
(#43467315)
I'm not qualified to offer an accurate eli5, just a hand-wavy explanation...
Category theory can be leveraged to make faster theorem provers (making complex symbolic reasoning practical at larger scales).
Don't ask me how, hopefully someone who studies it will chime in and correct me / expand.