arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2301.13867cs.LGcs.AIcs.CL
收藏

ChatGPT 的数学能力

Mathematical Capabilities of ChatGPT

University of Oxford(牛津大学);University of Vienna(维也纳大学);Institute for Advanced Study(普林斯顿高等研究院);University of Cambridge(剑桥大学);Vienna University of Technology(维也纳工业大学);VERSES Research Lab(VERSES研究实验室)

Simon Frieder, Luca Pinchetti, Alexis Chevalier, Ryan-Rhys Griffiths, Tommaso Salvatori, Thomas Lukasiewicz, Philipp Christian Petersen, Julius Berner

AI总结:

本文构建由数学研究者整理、覆盖研究生层次并区分推理维度的 GHOSTS 与 miniGHOSTS 数据集,系统评测 ChatGPT 和 GPT-4,发现其可作事实查询助手但整体数学能力远低于研究生水平。

摘要:

我们采用一种新颖方法,通过在公开数据集和手工构建数据集上进行测试,考察了两个迭代版本的 ChatGPT(分别发布于2023年1月9日和2023年1月30日)以及 GPT-4 的数学能力。与存在大型形式化证明数据库(例如 Lean Mathematical Library)的形式化数学不同,当前用于基准测试语言模型的自然语言数学数据集要么只涵盖初等数学,要么规模很小。我们通过公开发布两个新数据集 GHOSTS 和 miniGHOSTS 来解决这一问题。它们是首批由在职数学研究者整理的自然语言数据集,其特点是:(1)旨在覆盖研究生层次的数学;(2)提供对语言模型数学能力的全面概览;(3)区分数学推理的多个维度。这些数据集还通过模拟数学家日常职业活动中出现的使用场景,测试 ChatGPT 和 GPT-4 是否能够成为专业数学家的有用助手。我们在一系列细粒度性能指标上对这些模型进行基准测试。对于高等数学而言,这是迄今为止最详尽的评估工作。我们发现,ChatGPT 作为查询事实的数学助手最为成功,可充当数学搜索引擎和知识库接口。GPT-4 此外还可用于本科层次的数学,但在研究生层次的难度上会失败。与媒体中许多关于 GPT-4 和 ChatGPT 解题考试能力的正面报道相反(这可能是选择偏差的一个案例),它们的整体数学表现远低于研究生水平。因此,如果你的目标是利用 ChatGPT 通过研究生数学考试,那还不如抄袭普通同学的答案!

英文摘要:

We investigate the mathematical capabilities of two iterations of ChatGPT (released 9-January-2023 and 30-January-2023) and of GPT-4 by testing them on publicly available datasets, as well as hand-crafted ones, using a novel methodology. In contrast to formal mathematics, where large databases of formal proofs are available (e.g., the Lean Mathematical Library), current datasets of natural-language mathematics, used to benchmark language models, either cover only elementary mathematics or are very small. We address this by publicly releasing two new datasets: GHOSTS and miniGHOSTS. These are the first natural-language datasets curated by working researchers in mathematics that (1) aim to cover graduate-level mathematics, (2) provide a holistic overview of the mathematical capabilities of language models, and (3) distinguish multiple dimensions of mathematical reasoning. These datasets also test whether ChatGPT and GPT-4 can be helpful assistants to professional mathematicians by emulating use cases that arise in the daily professional activities of mathematicians. We benchmark the models on a range of fine-grained performance metrics. For advanced mathematics, this is the most detailed evaluation effort to date. We find that ChatGPT can be used most successfully as a mathematical assistant for querying facts, acting as a mathematical search engine and knowledge base interface. GPT-4 can additionally be used for undergraduate-level mathematics but fails on graduate-level difficulty. Contrary to many positive reports in the media about GPT-4 and ChatGPT's exam-solving abilities (a potential case of selection bias), their overall mathematical performance is well below the level of a graduate student. Hence, if your goal is to use ChatGPT to pass a graduate-level math exam, you would be better off copying from your average peer!

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑