Paper on Peng Tan's AI Bloghttps://c44db530.hobbytp-github-io.pages.dev/zh/tags/paper/一个关注 AI 各领域的专题博客MIT RLM(递归语言模型)-克服LLM上下文窗口限制和性能衰减https://c44db530.hobbytp-github-io.pages.dev/zh/papers/rlm/Tue, 20 Jan 2026 22:18:36 +0800https://c44db530.hobbytp-github-io.pages.dev/zh/papers/rlm/MIT 递归语言模型（RLM）模型，旨在攻克大语言模型在处理超长文本时的上下文衰减与窗口限制难题。A Practical Guide for Designing, Developing, and Deploying Production-Grade Agentic AI Workflowshttps://c44db530.hobbytp-github-io.pages.dev/zh/mas/agentic_workflow_practices/Thu, 11 Dec 2025 20:10:00 +0800https://c44db530.hobbytp-github-io.pages.dev/zh/mas/agentic_workflow_practices/论文介绍了生产级智能体AI工作流的设计、开发与部署指南。通义千问（Qwen）系列模型概览https://c44db530.hobbytp-github-io.pages.dev/zh/qwen/qwen_overview/Sun, 07 Sep 2025 10:10:00 +0800https://c44db530.hobbytp-github-io.pages.dev/zh/qwen/qwen_overview/通义千问（Qwen）系列模型概览Qwen3 Tech Report解读https://c44db530.hobbytp-github-io.pages.dev/zh/qwen/qwen3/Tue, 13 May 2025 23:03:00 +0800https://c44db530.hobbytp-github-io.pages.dev/zh/qwen/qwen3/全方位解读Qwen3的论文技术报告