火星财经消息，据 OpenAI 官方博客，OpenAI 推出具备推理能力的人工智能模型 o1，内部代号为「草莓」。OpenAI o1 可以推理复杂的任务，并解决比之前科学、编码和数学模型更难的问题。在测试中 OpenAI o1 在物理、化学和生物学等具有挑战性的基准任务上的表现与博士生类似，其在数学和编码方面表现出色。在国际数学奥林匹克（IMO）资格考试中，GPT-4o 只正确解决了 13% 的问题，而推理模型的得分为 83%，OpenAI o1 的编码能力在 Codeforces 竞赛中达到了第 89 个百分点。
作为早期模型 OpenAI o1 尚不具备 ChatGPT 的许多有用功能，例如浏览网页以获取信息以及上传文件和图像，GPT-4o 在短期内将更加强大。但对于复杂的推理任务来说，这是一个重大进步，代表了人工智能能力的新水平。鉴于此将计数器重置为 1，并将该系列命名为 OpenAI o1。医疗保健研究人员可以使用 o1 来注释细胞测序数据，物理学家可以使用 o1 生成量子光学所需的复杂数学公式，所有领域的开发人员可以使用 o1 来构建和执行多步骤工作流程。
OpenAI 还发布了 OpenAI o1-mini，这是一种更快、更便宜的推理模型，在编码方面特别有效。作为较小的模型，o1-mini 比 o1-preview 便宜 80%，这使其成为需要推理但不需要广泛的世界知识的应用程序的强大且经济高效的模型。OpenAI 计划为所有 ChatGPT 免费用户提供 o1-mini 访问权限。

<div>火星财经消息，据 OpenAI 官方博客，OpenAI 推出具备推理能力的人工智能模型 o1，内部代号为「草莓」。OpenAI o1 可以推理复杂的任务，并解决比之前科学、编码和数学模型更难的问题。在测试中 OpenAI o1 在物理、化学和生物学等具有挑战性的基准任务上的表现与博士生类似，其在数学和编码方面表现出色。在国际数学奥林匹克（IMO）资格考试中，GPT-4o 只正确解决了 13% 的问题，而推理模型的得分为 83%，OpenAI o1 的编码能力在 Codeforces 竞赛中达到了第 89 个百分点。<p>作为早期模型 OpenAI o1 尚不具备 ChatGPT 的许多有用功能，例如浏览网页以获取信息以及上传文件和图像，GPT-4o 在短期内将更加强大。但对于复杂的推理任务来说，这是一个重大进步，代表了人工智能能力的新水平。鉴于此将计数器重置为 1，并将该系列命名为 OpenAI o1。医疗保健研究人员可以使用 o1 来注释细胞测序数据，物理学家可以使用 o1 生成量子光学所需的复杂数学公式，所有领域的开发人员可以使用 o1 来构建和执行多步骤工作流程。</p><p>OpenAI 还发布了 OpenAI o1-mini，这是一种更快、更便宜的推理模型，在编码方面特别有效。作为较小的模型，o1-mini 比 o1-preview 便宜 80%，这使其成为需要推理但不需要广泛的世界知识的应用程序的强大且经济高效的模型。OpenAI 计划为所有 ChatGPT 免费用户提供 o1-mini 访问权限。</p></div>

OpenAI推出具备推理能力的人工智能模型o1，在数学和编码方面表现卓越

According to the OpenAI official blog, OpenAI has released an artificial intelligence model called o1, codenamed "Strawberry", with reasoning capabilities. OpenAI o1 can reason about complex tasks and solve problems that are more challenging than previous scientific, coding, and mathematical models. In testing, the performance of OpenAI o1 on challenging benchmark tasks in physics, chemistry, and biology was similar to that of doctoral students, and it performed exceptionally well in mathematics and coding. On the International Mathematical Olympiad (IMO) qualifying exam, GPT-4o only solved 13% of the problems correctly, while the reasoning model scored 83%, and OpenAI o1's coding ability reached the 89th percentile in the Codeforces competition.

As an early model, OpenAI o1 does not yet have many of the useful features of ChatGPT, such as browsing the web for information and uploading files and images, and GPT-4o will be more powerful in the short term. However, this represents a significant advancement in the capabilities of artificial intelligence for complex reasoning tasks. Given this, the counter will be reset to 1, and this series will be named OpenAI o1.

Medical researchers can use o1 to annotate cell sequencing data, physicists can use o1 to generate the complex mathematical formulas required for quantum optics, and developers in all fields can use o1 to build and execute multi-step workflows.

OpenAI has also released OpenAI o1-mini, a faster and cheaper reasoning model that is particularly effective in coding. As a smaller model, o1-mini is 80% cheaper than o1-preview, making it a powerful and cost-effective model for applications that require reasoning but not extensive world knowledge. OpenAI plans to provide o1-mini access to all ChatGPT free users.

<div>According to the official blog of OpenAI, OpenAI has launched an artificial intelligence model o1 with reasoning capabilities, code-named &quot;Strawberry&quot;. OpenAI o1 can reason about complex tasks and solve more difficult problems than previous scientific, coding, and mathematical models. In the test, OpenAI o1 performed similarly to a doctoral student on challenging benchmark tasks such as physics, chemistry, and biology, and excelled in mathematics and coding. In the International Mathematical Olympiad (IMO) qualifying exam, GPT-4o only correctly solved 13% of the problems, while the reasoning model scored 83%. OpenAI o1&#39;s coding ability reached the 89th percentile in the Codeforces competition.<p> As an early model, OpenAI o1 does not yet have many useful features of ChatGPT, such as browsing the web for information and uploading files and images. GPT-4o will be more powerful in the short term. But for complex reasoning tasks, this is a major improvement and represents a new level of artificial intelligence capabilities. In view of this, the counter is reset to 1 and the series is named OpenAI o1. Healthcare researchers can use o1 to annotate cell sequencing data, physicists can use o1 to generate complex mathematical formulas required for quantum optics, and developers in all fields can use o1 to build and execute multi-step workflows.</p><p> OpenAI also released OpenAI o1-mini, a faster and cheaper inference model that is particularly effective at encoding. As a smaller model, o1-mini is 80% cheaper than o1-preview, making it a powerful and cost-effective model for applications that require inference but do not require extensive world knowledge. OpenAI plans to provide o1-mini access to all ChatGPT free users.</p></div>

OpenAI launches AI model o1 with reasoning capabilities, excelling in mathematics and coding

Author: The Economist
Compiled by: TechFlow TechFlow
---------------------------------------------------------------
TechFlow Dive: Despite Bitcoin's price remaining above $70,000, the crypto market is experiencing an unprecedented "lonely winter." This article delves into the differences between this decline and previous ones: the chain reaction of leveraged liquidations, the fact that once-promising ETFs have now become drivers of price drops, and the most...

Bitcoin fell below $70,000. How did the "atmosphere" in the crypto market disappear?

In 2026, the question facing investors will no longer be "whether to allocate assets," but rather "how much to allocate and what tools to use for allocation."

ARK Invest: The Institutionalization of Bitcoin

Altcoins are posting gains as the crypto market cap ticks up—but experts warn structural trend reversal remains elusive.