ARC Prize：GPT-6.1 Sol 在 ARC-AGI-3 上更高推理档位决策更优、推理成本更低
R to @OpenAI: Like GPT-6 Astra, GPT-6.1 Sol made better decisions at higher reasoning levels on ARC-AGI-3, requiring fewer actions to complete levels, which reduced the total inference cost. For example, on sk48 using the provider adapter (which preserves opaque reasoning state between turns and enables auto compaction), GPT-6.1 Sol with max reasoning worked out how to reposition the colored blocks around an obstacle sooner, while low spent much longer revisiting blocked moves and testing controls that didn't help. That helped it complete the game in 463 actions versus 1,078 with low reasoning. Watch the replay: https://arcprize.org/replay/c299de73-6d1c-45f5-be74-c3552c90af73
ARC Prize 称 GPT-6.1 Sol 在 ARC-AGI-3 上提高推理档位后决策更优，完成关卡所需动作更少，从而降低了总推理成本。在 sk48 上，使用 provider adapter（保留跨轮不透明推理状态并支持自动压缩）时，max reasoning 的 GPT-6.1 Sol 以 463 个动作完成游戏，而 low reasoning 用了 1,078 个。
来源：X：ARC Prize (@arcprize) · x.lingyaoai.com