收录解读
- 分级:`颠覆性` - 正式标题:`AlphaEvolve: A coding agent for scientific and algorithmic discovery` - 原文:`2025-06-17-S1_AlphaEvolve-AlphaEvolve_A_coding_agent_for_scientific_and_algorithmic_discovery.pdf` - 抽取:`extracted.md`
## 重写摘要
AlphaEvolve 的关键突破,不是“让 LLM 写代码”,而是让 LLM 在可执行、可评测的进化闭环里持续改进算法。系统把候选程序交给评估器打分,再用进化式搜索不断保留高分变体,从而把语言模型的创造性和程序执行的确定性绑定在一起。结果是它不只会提出想法,而是能在反馈回路里发现更优算法。
这篇白皮书最重要的证据来自两类场景。第一类是科学与数学:AlphaEvolve 找到了两个 4×4 复矩阵只用 48 次标量乘法的程序,并在高维组合与几何构造问题上刷新记录。第二类是现实基础设施:它为 Google 数据中心调度发现了更优启发式,为 FlashAttention 相关内核和 TPU 电路设计给出可落地优化,并报告 Borg 调度长期回收约 0.7% 的全局算力资源。
## 为什么重要
它把“可验证输出”变成智能复利的核心。只要程序可以被执行和打分,模型就不再只是一次性生成器,而能成为持续发现算法和优化基础设施的搜索系统。这对自动科学发现、算法设计和 AI 自我改进都很关键。
## 局限
当前公开载体是 DeepMind 白皮书,发布时间为 `2025-06-17`,不在你最初的三个月窗口内。它的很多结果依赖专有评估环境和 Google 内部基础设施,外部复现门槛仍然很高。
原始摘要与中文对照
中文对照翻译
在这篇白皮书中,我们介绍了AlphaEvolve,一个进化式编码智能体,它显著增强了最先进LLM在极具挑战性的任务上的能力,例如解决开放性科学问题或优化计算基础设施的关键部分。AlphaEvolve协调了一个由LLM组成的自主流水线,其任务是通过直接修改代码来改进算法。采用进化方法,持续接收来自一个或多个评估器的反馈,AlphaEvolve迭代地改进算法,有可能带来新的科学和实践发现。我们通过将其应用于许多重要的计算问题,展示了该方法的广泛适用性。当应用于优化Google大规模计算堆栈的关键组件时,AlphaEvolve开发了一种更高效的数据中心调度算法,在硬件加速器的电路设计中发现了一个功能等效的简化,并加速了支撑AlphaEvolve本身的LLM的训练。此外,AlphaEvolve发现了新颖的、可证明正确的算法,这些算法在数学和计算机科学的一系列问题上超越了最先进的解决方案,显著扩展了先前自动化发现方法的范围 (Romera-Paredes et al., 2023)。值得注意的是,AlphaEvolve开发了一种搜索算法,该算法找到了一种使用48次标量乘法来乘以两个4 × 4复值矩阵的程序;在这种情况下,这是56年来首次超越Strassen算法的改进。我们相信AlphaEvolve及其类似的编码智能体可以在改进科学和计算领域许多问题的解决方案方面产生重大影响。
原始摘要
In this white paper, we present AlphaEvolve, an evolutionary coding agent that substantially enhances capabilities of state-of-the-art LLMs on highly challenging tasks such as tackling open scientific problems or optimizing critical pieces of computational infrastructure. AlphaEvolve orchestrates an autonomous pipeline of LLMs, whose task is to improve an algorithm by making direct changes to the code. Using an evolutionary approach, continuously receiving feedback from one or more evaluators, AlphaEvolve iteratively improves the algorithm, potentially leading to new scientific and practical discoveries. We demonstrate the broad applicability of this approach by applying it to a number of important com- putational problems. When applied to optimizing critical components of large-scale computational stacks at Google, AlphaEvolve developed a more efficient scheduling algorithm for data centers, found a functionally equivalent simplification in the circuit design of hardware accelerators, and acceler- ated the training of the LLM underpinning AlphaEvolve itself. Furthermore, AlphaEvolve discovered novel, provably correct algorithms that surpass state-of-the-art solutions on a spectrum of problems in mathematics and computer science, significantly expanding the scope of prior automated discovery methods (Romera-Paredes et al., 2023). Notably, AlphaEvolve developed a search algorithm that found a procedure to multiply two 4 × 4 complex-valued matrices using 48 scalar multiplications; offering the first improvement, after 56 years, over Strassen’s algorithm in this setting. We believe AlphaEvolve and coding agents like it can have a significant impact in improving solutions of problems across many areas of science and computation.