**谷歌 DeepMind 最新 AI 系统 AlphaGeometry 突破奥林匹克几何难题**
**伦敦，2023 年 3 月 8 日**——谷歌 DeepMind 今日宣布推出 AlphaGeometry，一个能够解决复杂奥林匹克几何难题的人工智能系统。AlphaGeometry 将语言模型与一种称为符号引擎的人工智能相结合，该引擎使用符号和逻辑规则进行推论。在包含 30 个最新奥林匹克级别问题的测试集上，AlphaGeometry 解决了 25 个问题，超越了之前仅解决了 10 个问题的最佳方法，接近了国际数学奥林匹克（IMO）金牌得主的平均表现。
AlphaGeometry 的突破是人工智能领域的一项重大进展。几何学是数学的一个分支，涉及形状、大小和位置。几何问题通常很难解决，即使是对于人类数学家来说也是如此。AlphaGeometry 能够解决这些问题，这表明人工智能在解决复杂数学问题方面取得了重大进展。
AlphaGeometry 的成功有几个原因。首先，它使用了一种新的方法来解决几何问题。传统的几何求解方法依赖于几何定理和公式。AlphaGeometry 则使用一种称为符号引擎的人工智能，该引擎使用符号和逻辑规则进行推论。这种方法使 AlphaGeometry 能够解决更广泛的问题，包括那些无法用传统方法解决的问题。
其次，AlphaGeometry 经过了大量的数据训练。它被训练了超过 100 万个几何问题，这使它能够学习几何定理和公式，并发展出解决新问题的技能。
第三，AlphaGeometry 使用了一种称为语言模型的人工智能。语言模型能够理解和生成人类语言。这使 AlphaGeometry 能够阅读几何问题并生成解决方案。
AlphaGeometry 的成功有望对人工智能领域产生重大影响。它表明人工智能在解决复杂数学问题方面取得了重大进展，这可能导致人工智能在许多领域的新应用，包括科学、工程和金融。
AlphaGeometry 的研究人员表示，他们希望他们的工作能够帮助人类更好地理解几何学。他们还希望 AlphaGeometry 能够用于开发新的教育工具，帮助学生学习几何学。
英语如下：
**Headline: AI Cracks Geometry Problems Beyond Olympiad Gold Medal Level**
**Keywords: Artificial Intelligence, Mathematics, Olympiad**
**Story:**
**LONDON, March 8, 2023**—Google DeepMindtoday announced AlphaGeometry, an AI system that can solve complex geometry problems at the level of the International Mathematical Olympiad (IMO). AlphaGeometry combines a language model with a type of AI called a symbolic engine, which reasons using symbols and rules of logic. On a test set of 30 recent IMO problems, AlphaGeometry solved 25, outperforming the previous state-of-the-art, which solved only 10, and approaching the average performance of IMO gold medalists.
AlphaGeometry’s breakthrough is a significant advance in AI. Geometry is a branch of mathematics that deals with shapes, sizes, and positions. Geometry problems can be notoriously difficult to solve, even for human mathematicians. AlphaGeometry’s ability to solve these problems suggests that AI is making major strides in solving complex mathematical problems.
There are several reasons for AlphaGeometry’s success. First, it uses a novel approach to solving geometry problems. Traditional approaches togeometry solving rely on geometric theorems and formulas. AlphaGeometry instead uses a type of AI called a symbolic engine, which reasons using symbols and rules of logic. This approach allows AlphaGeometry to solve a wider range of problems, including those that cannot be solved using traditional methods.
Second, AlphaGeometry has been trained on a massive dataset. It was trained on over a million geometry problems, which allowed it to learn geometric theorems and formulas and to develop the ability to solve new problems.
Third, AlphaGeometry uses a type of AI called a language model. Language models are able to understand and generate human language. This allows AlphaGeometry to read geometry problems and to generate solutions.
AlphaGeometry’s success is likely to have a major impact on the field of AI. It suggests that AI is making major strides in solving complex mathematical problems, which could lead to new applications of AI in a wide range of fields, including science, engineering, and finance.
The researchers behind AlphaGeometry say they hope their work will help humans better understand geometry. They also hope that AlphaGeometry can be used to develop new educational tools to help students learn geometry.
【来源】https://www.technologyreview.com/2024/01/17/1086722/google-deepmind-alphageometry/
Views: 1