2210, 2026 expand vs repeat: Shared Storage and Summed Gradients 黎 浩然/ 22 10 月, 2026/ 大语言模型/LARGELANGUAGEMODEL/LLM, 教程/TUTORIAL, 计算机/COMPUTER/ 0 comments Read More
2210, 2026 expand与repeat：省复制，为何梯度仍相加 黎 浩然/ 22 10 月, 2026/ 大语言模型/LARGELANGUAGEMODEL/LLM, 教程/TUTORIAL, 计算机/COMPUTER/ 0 comments Read More
2210, 2026 GitHub Read Replicas: Cached Objects Can Still Leave Stale References 黎 浩然/ 22 10 月, 2026/ 大语言模型/LARGELANGUAGEMODEL/LLM, 教程/TUTORIAL, 计算机/COMPUTER/ 0 comments Read More
2210, 2026 GitHub读副本：对象已缓存，引用为何仍过期 黎 浩然/ 22 10 月, 2026/ 大语言模型/LARGELANGUAGEMODEL/LLM, 教程/TUTORIAL, 计算机/COMPUTER/ 0 comments Read More
2210, 2026 PBC Weather Forecasts: Valid Probabilities Are Not Enough 黎 浩然/ 22 10 月, 2026/ 大语言模型/LARGELANGUAGEMODEL/LLM, 教程/TUTORIAL, 计算机/COMPUTER/ 0 comments Read More
2210, 2026 PBC天气预报解读：概率有效，不等于预报可靠 黎 浩然/ 22 10 月, 2026/ 大语言模型/LARGELANGUAGEMODEL/LLM, 教程/TUTORIAL, 计算机/COMPUTER/ 0 comments Read More
2210, 2026 GitHub Per-Reference Coordination: Why Agent Writes Need Not Share One Queue 黎 浩然/ 22 10 月, 2026/ 大语言模型/LARGELANGUAGEMODEL/LLM, 教程/TUTORIAL, 计算机/COMPUTER/ 0 comments Read More
2210, 2026 GitHub逐引用协调：多代理写入为何不必全排队 黎 浩然/ 22 10 月, 2026/ 大语言模型/LARGELANGUAGEMODEL/LLM, 教程/TUTORIAL, 计算机/COMPUTER/ 0 comments Read More
2110, 2026 torch.where and NaN Gradients: Why Masking Outputs Can Be Too Late 黎 浩然/ 21 10 月, 2026/ 大语言模型/LARGELANGUAGEMODEL/LLM, 教程/TUTORIAL, 计算机/COMPUTER/ 0 comments Read More
2110, 2026 torch.where：输出正常，梯度为何仍是NaN 黎 浩然/ 21 10 月, 2026/ 大语言模型/LARGELANGUAGEMODEL/LLM, 教程/TUTORIAL, 计算机/COMPUTER/ 0 comments Read More