NVIDIA编码智能体AVO在ARC-AGI-3基准测试中满分通过
Holy: NVIDIA’s coding agent AVO scored 100% on ARC-AGI-3’s 25 public games, solv…
Holy: NVIDIA’s coding agent AVO scored 100% on ARC-AGI-3’s 25 public games, solving all 183 levels.
Holy：NVIDIA 的编码代理 AVO 在 ARC-AGI-3 的 25 个公开游戏中取得 100% 的成绩，解决了全部 183 个关卡。
The agent receives no rules or stated goals. It must learn by trying things, observing the results and correcting its mistakes.
该代理未接收任何规则或明确目标。它必须通过尝试、观察结果并纠正错误来学习。
AVO succeeds by remembering what it learned and building on it over long periods instead of starting over when the model’s context resets.
AVO 的成功在于它能记住所学内容并在长期内在此基础上构建，而不是在模型上下文重置时重新开始。
Powered by Claude Opus 5, the same system previously worked autonomously for seven days optimizing GPU code.
由 Claude Opus 5 驱动，同一系统此前曾自主运行七天以优化 GPU 代码。
Really really cool to see what NVIDIA achieved here
看到 NVIDIA 在此取得的成就真的非常酷
更进一步：量化金融体系
看懂新闻只是起点——沿量化金融路径，把它变成能交付的工程能力