Smeddum's AGI Database
What it is
A neuro-symbolic inference engine designed for the ARC-AGI-2 competition. It uses a materialist architecture that employs topological auditing, persistent homology, and recursive decomposition to solve grid-based abstractions through symbolic reasoning rather than probabilistic weights.
The problem it solves
Overcomes the limitations of LLMs in handling complex spatial reasoning and abstract pattern recognition by using symbolic inference and topological invariants.
What you could use it for
- Solving ARC-AGI (Abstraction and Reasoning Corpus) tasks
- Symbolic reasoning and geometric manifold analysis
- Recursive decomposition of complex spatial problems
Read the source ↗Back to the library →
Summary written by an AI model from the source; check the source before you rely on it.