---
id: 20260701-T0-15
title: "几何求解新范式：Solver驱动自动形式化与定理提出"
title_en: "Geometry Solving: Solver-Driven Autoformalization and Theorem Proposing"
url: https://ai.daily.yangsir.net/daily/20260701-T0-15
issue_date: 2026-07-01
publish_date: 2026-06-30T04:00:00.000Z
category: research
source_name: "arXiv cs.AI"
source_url: https://arxiv.org/abs/2606.27926
---

# 几何求解新范式：Solver驱动自动形式化与定理提出

针对几何问题求解中的神经-符号瓶颈，新研究提出了Solver驱动的方法。该方法通过改进自动形式化处理多模态转换，并提出新的定理生成机制，提升了AI解决几何问题的能力。

## English Version

**Geometry Solving: Solver-Driven Autoformalization and Theorem Proposing**

This paper addresses bottlenecks in geometry problem solving, specifically in autoformalization and theorem proposing. It introduces a solver-driven framework to improve the neuro-symbolic paradigm for geometry tasks.

---

**来源**：[arXiv cs.AI](https://arxiv.org/abs/2606.27926)

**详情页**：https://ai.daily.yangsir.net/daily/20260701-T0-15

---

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