Qwenputer 探索將運算機制直接嵌入 LLM 權重,無須重新訓練原始語言模型;這個由 @divagr1925 開發的專案,已從在 Qwen3.5 中加入 RISC-V CPU,擴充套件為一系列模型。
AI 伴隨問答
針對這篇報導有疑問?可直接點選快捷問題或向 AI 追問背景脈絡與技術細節:
Qwenputer 系列將運算單元嵌入模型權重,並從 RISC-V CPU 延伸至檔案儲存與向量運算,呈現擴充模型能力的不同設計路徑。
Qwenputer explores a fascinating idea: embedding computational machinery directly into LLM weights, without retraining the original language model.
Developed by @divagr1925 , the project has evolved from a RISC-V CPU inside Qwen3.5 into a family of models exploring new ways to extend LLM capabilities. The full model family and an interactive demo are now available on ModelScope:
🚀 Explore all models:
https://t.co/uw47cLROuq
🧪 Try the interactive demo:
https://t.co/oo6RxHWlV5
Here's a closer look at the model family:
• Qwenputer embeds a functional RISC-V CPU inside Qwen3.5, using Gated DeltaNet's recurrent state as RAM.
• Qwenputer-Filesystem introduces persistent file storage within the model's recurrent state, enabling memory across sessions.
• Qwenputer-NPU extends the internal CPU with neural network instructions, allowing it to execute small models.
• QwenPU takes things further with a dedicated vector unit capable of running Karpathy's 260K-parameter Llama inside Qwen itself.
Original project 👉: https://t.co/ly2tHgBfoa