ARC Prize Finds Power Law: V2 Scores 98% Predictable from V1
ARC Prize analysis reveals a power law (v2 = v1^4.53) explaining about 98% of ARC-AGI-2 score variance, as researchers discuss ceiling effects on v1 and whether v2 extends discrimination among frontier models.
2026-10-29 ~ 2026-10-29 · 3 related posts
- ARC Prize finds v2 = v1^4.53 power curve explains 98% of ARC-AGI-2 score variance — GregKamradt · 2026-10-29
- ARC-AGI scores follow power law v2 = v1^4.53, explaining 98% of variance — typewriters · 2026-10-29
- ARC team: v1 hits ceiling effect as v2 extends headroom for frontier models — MindsAI_Jack · 2026-10-29