ARC-AGI-3 Re/Im Agent Spectator

Im side = bird's eye hypothesis (which transformation?) | Re side = exact location (which cells to click?)

🧠 = analytic solver (Im picks hypothesis → Re pins cell → ACTION6 click) 🎲 = CNN fallback (when no hypothesis clears the confidence threshold)

ARC API key

Game

Max steps

20 500

Fetch games → select → Watch

Current frame (🔴=wrong cells ⭐=target click)

Im side — hypothesis ranking

Im candidate — what the answer should look like

Action frequency

Gabor s-plane — cross terms (σ>0, ω>0)

Reward 🟡+50 WIN 🟡+10 level 🟢+0.1 change 🔴-0.01 dead

Re/Im duality in action: The Im side reads the whole board at once — symmetry maps, boundary contour, directional flow — and ranks candidate transformations by confidence. The Re side then diffs the current frame against the winning candidate and finds the exact cell (boundary-first, following Cauchy's principle) that most needs fixing. The agent emits ACTION6 at those precise coordinates instead of guessing randomly. CNN fires only when no analytic hypothesis clears 0.40 confidence.