理解推理 LLMs

This article describes the four main approaches to building reasoning models, or how we can enhance LLMs with reasoning capabilities. I hope this provides valuable insights and helps you navigate the rapidly evolving literature and hype surrounding this topic.

本文描述了构建推理模型的四种主要方法，或者说我们如何增强LLM的推理能力。我希望这能提供有价值的见解，并帮助您在快速发展的文献和围绕这一主题的炒作中导航。

In 2024, the LLM field saw increasing specialization. Beyond pre-training and fine-tuning, we witnessed the rise of specialized applications, from RAGs to code assistants. I expect this trend to accelerate in 2025, with an even greater emphasis on domain- and application-specific optimizations (i.e., "specializations").

在 2024 年，LLM 领域看到了越来越多的专业化。除了预训练和微调，我们见证了从 RAG 到代码助手的专业应用的兴起。我预计这一趋势将在 2025 年加速，进一步强调领域和应用特定的优化（即“专业化”）。

[

[

](https://substackcdn.com/image/fetch/f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd6ebc5c9-461f-4d3a-889b-b8ea4e14e5ba_1600x830.png)

](https://substackcdn.com/image/fetch/f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd6ebc5c9-461f-4d3a-889b-b8ea4e14e5ba_1600x830.png)

Stages 1-3 are the common steps to developing LLMs. Stage 4 specializes LLMs for specific use cases.

阶段1-3是开发LLMs的常见步骤。阶段4专门针对特定用例优化LLMs。

The development of reasoning models is one of these specializations. This means we refine LLMs to excel at complex tasks that are best solved with intermediate steps, such as puzzles, advanced math, and coding challenges. However, this specialization does not replace other LLM applications. Because transforming an LLM into a reasoning model also introduces certain drawbacks, which I will discuss later.

推理模型的发展是这些专业化之一。这意味着我们优化LLM以在复杂任务中表现出色，这些任务最好通过中间步骤解决，例如谜题、高级数学和编码挑战。然而，这种专业化并不取代其他LLM应用。因为将LLM转变为推理模型也会引入某些缺点，我稍后会讨论。

To give you a brief glimpse of what's covered below, in this article, I will:

为了让您简要了解下面的内容，在本文中，我将：

1. Explain the meaning of "reasoning model"

   解释“推理模型”的含义

2. Discuss the advantages and disadvantages of reasoning models

   讨...