<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/"><channel><title>论文 | AIM Blog</title><link>https://ai-mathematician.net/zh/publications/</link><description>论文的最新内容，来自AIM Blog</description><generator>Hugo</generator><language>zh</language><atom:link href="https://ai-mathematician.net/zh/publications/index.xml" rel="self" type="application/rss+xml"/><item><title>AI Mathematician: Towards Fully Automated Frontier Mathematical Research</title><link>https://arxiv.org/abs/2505.22451</link><pubDate>Wed, 28 May 2025 00:00:00 +0000</pubDate><guid>https://arxiv.org/abs/2505.22451</guid><category>论文</category><description>提出 AIM，一个面向前沿数学研究的大推理模型智能体框架，结合长程探索轨迹和面向验证的机制来处理研究级证明任务。</description></item><item><title>From Meta Idea to Advanced Mathematical Discovery -- Human-AI Co-Discovery of Sign-Embedding Quantum Algorithms</title><link>https://arxiv.org/abs/2606.24899</link><pubDate>Fri, 12 Jun 2026 00:00:00 +0000</pubDate><guid>https://arxiv.org/abs/2606.24899</guid><category>论文</category><description>呈现一个人机协同发现案例：AIM 支持路线探索、定理形成、推导生成与审计流程，并参与推动面向矩阵方程和矩阵函数的符号嵌入量子算法研究。</description></item><item><title>Pessimistic Verification for Open Ended Math Questions</title><link>https://arxiv.org/abs/2511.21522</link><pubDate>Wed, 26 Nov 2025 00:00:00 +0000</pubDate><guid>https://arxiv.org/abs/2511.21522</guid><category>论文</category><description>研究面向开放式数学证明的悲观验证流程：只要任一验证器发现关键错误，就拒绝该证明。</description></item><item><title>AI Mathematician as a Partner in Advancing Mathematical Discovery -- A Case Study in Homogenization Theory</title><link>https://arxiv.org/abs/2510.26380</link><pubDate>Thu, 30 Oct 2025 00:00:00 +0000</pubDate><guid>https://arxiv.org/abs/2510.26380</guid><category>论文</category><description>通过均匀化理论案例展示 AIM 辅助推理与人工定向干预如何支持完整数学证明的形成。</description></item><item><title>FormaRL: Enhancing Autoformalization with no Labeled Data</title><link>https://arxiv.org/abs/2508.18914</link><pubDate>Tue, 26 Aug 2025 00:00:00 +0000</pubDate><guid>https://arxiv.org/abs/2508.18914</guid><category>论文</category><description>提出 FormaRL，一个利用无标注数据、Lean 语法检查和大模型一致性检查提升自动形式化准确率的强化学习框架。</description></item></channel></rss>