📄️ 马斯克发布的Grok1 - 314B参数模型
This article discusses the release of Grok1 by Elon Musk's xAI, a 314 billion parameter open-source model known for its humor and interactivity.
📄️ 一文读懂 OpenAI超级对齐团队首个成果 - 用GPT2监督GPT-4
This page summarizes OpenAI's latest research on using a weaker model, GPT-2, to enhance the performance of a stronger model, GPT-4, as part of their superalignment initiative.
📄️ 史上最快大模型 Groq
Explore Groq, the new large model offering super-fast response speeds based on the open-source Mixtral 8x7B-32k model, boasting speeds of 500 tokens per second!
📄️ 我花了一周时间体验GeminiPro 1.5—太棒了
This page summarizes the experiences and significant advancements of the Gemini Pro 1.5 language model.
📄️ Claude3 深夜“奇袭” GPT
This page details the launch of the Claude3 family and compares its features and performance to GPT-4.
📄️ Sora 和之前 Runway 那些在架构上有啥区别？
This article explores the architectural differences between Sora and previous models like Runway.
📄️ 中学生能看懂 - Sora 原理解读
An in-depth look at Sora, OpenAI's powerful video generation AI, including its principles, technology, and limitations.
📄️ GPT5 伪装 gpt2-chatbot 神秘上线？
This page compares the mysterious gpt2-chatbot with GPT4, exploring its reasoning, math skills, coding ability, and ASCII art capabilities.
📄️ 一手测评GPT4o十项能力 - 中文场景下轻松超过GPT4
An in-depth assessment of GPT-4o's performance across ten key capabilities in the Chinese context, demonstrating its superiority over GPT-4.
📄️ 谷歌对OpenAI的回击 - 发布会上透露了什么？
This page provides a detailed analysis of Google's recent counter to OpenAI's advancements, including new projects, models, and tools.