Shadow Walker | 松烟阁的 twitter 的 RSS 预览

so GPT-4.5 is not a RLM, what is the magic to be a thoughtful person?

2025-02-28 08:25:17

so GPT-4.5 is not a RLM, what is the magic to be a thoughtful person?

Sam Altman: GPT-4.5 is ready!

good news: it is the first model that feels like talking to a thoughtful person to me. i have had several moments where i've sat back in my chair and been astonished at getting actually good advice from an AI.

bad news: it is a giant, expensive model. we

Re @yihong0618 精辟

2025-02-28 08:21:07

Re @yihong0618 精辟

研究一下DeepSeek涉及的推理机制（Reasoning Schema），包括其定义、组成部分（推理结构、策略和操作）以及工作原理。通过一个24点游戏的例子展示DeepSeek的思考...

2025-02-24 16:14:45

研究一下DeepSeek涉及的推理机制（Reasoning Schema），包括其定义、组成部分（推理结构、策略和操作）以及工作原理。通过一个24点游戏的例子展示DeepSeek的思考过程，并对比了推理（Reason）与推断（Inference）的区别，强调了推理在逻辑性和解释性方面的优势
https://www.edony.ink/guan-yu-deepseekwo-shi-zen-yao-yan-jiu-de-3-2/

Re @ghczyy17 又休假啦～

2025-02-22 19:47:07

Re @ghczyy17 又休假啦～

oh, 感觉scaling law又要有头皮发麻了

2025-02-18 16:19:26

oh, 感觉scaling law又要有头皮发麻了

DeepSeek: 🚀 Introducing NSA: A Hardware-Aligned and Natively Trainable Sparse Attention mechanism for ultra-fast long-context training & inference!

Core components of NSA:
• Dynamic hierarchical sparse strategy
• Coarse-grained token compression
• Fine-grained token selection

💡 With

Re @ghczyy17 你的好多🥳

2025-02-18 15:53:43

Re @ghczyy17 你的好多🥳

Shadow Walker | 松烟阁修改

Shadow Walker | 松烟阁的 twitter 的 RSS 预览

该作者的社会化媒体