---
id: 20260916-T0-03
title: "ZGCM-1：完全开放的高效数学与Agent搜索基础模型"
title_en: "ZGCM-1: A Fully Open 7B Model for Math and Agentic Search"
url: https://ai.daily.yangsir.net/daily/20260916-T0-03
issue_date: 2026-09-16
publish_date: 2026-09-15T04:00:00.000Z
category: research
source_name: "arXiv cs.AI"
source_url: https://arxiv.org/abs/2609.13356
---

# ZGCM-1：完全开放的高效数学与Agent搜索基础模型

研究人员发布 ZGCM-1，一个完全开放的 7B 密集基础模型，从零开始训练，在数据、系统和算法效率上做了极致优化。核心前提是：紧凑模型无法被动记忆整个开放网络，但可以通过过度训练获得特定能力。ZGCM-1 面向数学和 Agent 搜索任务，全部开源。对资源有限的研究团队来说，这提供了一个可复现的高效基础模型选项。

## English Version

**ZGCM-1: A Fully Open 7B Model for Math and Agentic Search**

Researchers introduced ZGCM-1, a fully open 7B dense foundation model trained from scratch with extreme data, system, and algorithmic efficiency. The core premise: compact models cannot passively memorize the open web, but can be overtrained for specific capabilities. ZGCM-1 targets math and agentic search tasks and is fully open-source. For resource-constrained research teams, it offers a reproducible and efficient foundation model option.

---

**来源**：[arXiv cs.AI](https://arxiv.org/abs/2609.13356)

**详情页**：https://ai.daily.yangsir.net/daily/20260916-T0-03

---

*智语观潮 · Daily — https://ai.daily.yangsir.net/llms.txt*