Hacker News 热门(buzzing.cc 中文翻译)
57AI 编辑部评分,满分 100

Mistral 获"基于代码实现的工具调用"专利,涵盖沙箱执行与暂停恢复机制

2026-08-11 00:11· 2小时前· theanonymousone
AI 导读

Mistral AI 获得美国专利 US 12,670,045 B1,涵盖一种基于代码实现的工具调用方法。该方法由大语言模型(LLM)生成封装工具调用的代码块,在服务器沙箱中执行,遇待处理调用时暂停并将请求发送至客户端执行,随后恢复代码块运行并替换结果。专利共 20 项权利要求,发明人为 Gabriel Vergnaud。

US 12,670,045 B1
Code implemented tool calls
Gabriel Vergnaud, Paris (FR)
Assigned to Mistral AI, Paris (FR)
Filed by Mistral AI, Paris (FR)
Filed on Mar. 4, 2026, as Appl. No. 19/557,103.
Int. Cl. G06F 9/54 (2006.01)
CPC G06F 9/547 (2013.01) 20 Claims
OG exemplary drawing
 
1. A method, comprising:
receiving, at a server, a user request for execution of one or more tool calls;
generating, by a large language model (LLM), a code block in a programming language, the code block configured to encapsulate the one or more tool calls;
executing, by the server, the code block in a sandbox;
in response to obtaining a pending tool call, pausing execution of the code block;
transmitting the pending tool call to a client for execution;
receiving, from the client, a first result of the pending tool call;
resuming execution of the code block and substituting the first result of the pending tool call for the pending tool call in the code block; and
returning a second result of the executed code block to the LLM.

来源:Hacker News 热门(buzzing.cc 中文翻译) · patentsgazette.uspto.gov

Mistral 获"基于代码实现的工具调用"专利,涵盖沙箱执行与暂停恢复机制

Hacker News 热门(buzzing.cc 中文翻译)·2026-08-11 00:11·2小时前·theanonymousone
AI 导读

Mistral AI 获得美国专利 US 12,670,045 B1,涵盖一种基于代码实现的工具调用方法。该方法由大语言模型(LLM)生成封装工具调用的代码块,在服务器沙箱中执行,遇待处理调用时暂停并将请求发送至客户端执行,随后恢复代码块运行并替换结果。专利共 20 项权利要求,发明人为 Gabriel Vergnaud。

原文 · 保持原样,未翻译
US 12,670,045 B1
Code implemented tool calls
Gabriel Vergnaud, Paris (FR)
Assigned to Mistral AI, Paris (FR)
Filed by Mistral AI, Paris (FR)
Filed on Mar. 4, 2026, as Appl. No. 19/557,103.
Int. Cl. G06F 9/54 (2006.01)
CPC G06F 9/547 (2013.01) 20 Claims
OG exemplary drawing
 
1. A method, comprising:
receiving, at a server, a user request for execution of one or more tool calls;
generating, by a large language model (LLM), a code block in a programming language, the code block configured to encapsulate the one or more tool calls;
executing, by the server, the code block in a sandbox;
in response to obtaining a pending tool call, pausing execution of the code block;
transmitting the pending tool call to a client for execution;
receiving, from the client, a first result of the pending tool call;
resuming execution of the code block and substituting the first result of the pending tool call for the pending tool call in the code block; and
returning a second result of the executed code block to the LLM.

来源:Hacker News 热门(buzzing.cc 中文翻译)· patentsgazette.uspto.gov