ARC-AGI-3 Graph Exploration Agent
What it is
A graph-based reinforcement learning approach for the ARC-AGI-3 competition. It uses Breadth-First Search (BFS) to explore a state graph, mapping grid configurations as nodes and actions as edges to find optimal solutions.
The problem it solves
Overcomes the limitations of standard LLM/CNN approaches in ARC tasks by using a structured state-space search and a 'Click Solver' for coordinate-based actions.
What you could use it for
- Solving ARC-AGI (Abstraction and Reasoning Corpus) tasks
- Algorithmic state-space exploration
- Pathfinding in grid-based environments using graph theory
Read the source ↗Back to the library →
Summary written by an AI model from the source; check the source before you rely on it.