Anthropic：Claude在科学博客展示九圈散射振幅计算能力
New on the Science Blog: Yes, Claude can do Nine Loops.
Claude在复杂科学推理上的新突破，展示了其在特定领域超越人类专家的能力，值得关注。
New on the Science Blog: Yes, Claude can do Nine Loops.
科学博客新帖：是的，Claude 能完成九圈计算。
Theoretical physicists predict how particles behave using formulas called scattering amplitudes. These are notoriously hard to compute, so researchers work with layers of increasingly fine corrections called “loops”—each added loop makes the answer more precise but takes exponentially more computation. Most calculations stop at two or three loops. Eight loops was the previous record in a simplified model physicists use as a testing ground (planar N=4 super-Yang-Mills), set by SLAC's Lance Dixon and collaborators.
理论物理学家使用称为散射振幅的公式来预测粒子的行为。这些公式以难以计算著称，因此研究人员采用逐层叠加越来越精细的修正项的方法，这些修正项被称为“圈”——每增加一圈会使答案更精确，但所需的计算量呈指数级增长。大多数计算在二圈或三圈处停止。此前在物理学家用作测试平台的简化模型（平面 N=4 超杨-米尔斯理论）中，史丹福直线加速器中心（SLAC）的兰斯·迪克森（Lance Dixon）及其合作者创下的纪录是八圈。
Last month, physicist and science writer @4gravitons issued a challenge: could an AI push past eight loops in this model, using only the compute budget an academic could reasonably access?
上个月，物理学家兼科学作家 @4gravitons 发起了一项挑战：能否让 AI 在该模型中突破八圈的界限，且仅使用学术界可合理获取的计算资源？
Given a single prompt describing the nine-loop problem, Claude ran largely unsupervised for days in Claude Science and solved it using methods developed by Dixon and his colleagues, at a total cost of a few thousand dollars. Dixon independently verified the result, and von Hippel wrote about the experience for our blog.
在接收到描述九圈问题的单个提示后，Claude 在 Claude Science 中基本自主运行了数天，并运用迪克森及其同事开发的方法解决了该问题，总成本仅为数千美元。迪克森独立验证了结果，冯·希佩尔（von Hippel）则为我们的博客撰写了此次体验的相关文章。
Read more: https://www.anthropic.com/research/yes-claude-can-do-nine-loops
阅读更多：https://www.anthropic.com/research/yes-claude-can-do-nine-loops
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