ARCANA: A Reflective Multi-Agent Program Synthesis Framework for ARC-AGI-2 Reasoning
For researchers tackling abstract reasoning tasks with limited test-time compute, ARCANA offers a structured multi-agent approach that outperforms single-agent methods.
ARCANA introduces a multi-agent framework that decomposes ARC-AGI-2 tasks into perception, hypothesis generation, execution, and reflection, achieving improved reasoning efficiency and solution quality under strict constraints.
We present ARCANA, a collaborative multi agent framework for solving ARC AGI 2 tasks under strict test time and hardware constraints. ARCANA decomposes each task into iterative perception, hypothesis generation, symbolic execution, and reflective refinement. A perceptual grounding agent builds object centric scene graphs from raw grids, a latent program policy proposes diverse DSL programs, a symbolic executor verifies candidates on demonstrations, and a reflective agent synthesizes failure driven feedback for the next turn. These agents communicate through a shared differentiable blackboard and are scheduled by a learned meta controller. The design combines structured program search with adaptive multi turn correction, improving reasoning efficiency and solution quality on challenging abstract transformation tasks.