AI“短板”暴露：研究发现GPT-4 Turbo回答高级历史题准确率仅46%
研究结果于上月在知名 AI 会议 NeurIPS 上公布，结果显示， 即使是表现最佳的 GPT-4 Turbo 模型，其准确率也仅为 46%，并不比随机猜测高多少。 论文合著者、伦敦大学学院计算机科学副教授 Maria del Rio-Chanona 表示：“这项研究的主要结论是，尽管 LLMs ...
Are ChatGPT Plus or Pro worth it? Here's how they compare to the free version
ChatGPT Pro is 10 times the price of ChatGPT Plus. Is either worth the money or should you stick to the free version? Here's ...
腾讯网25 天
重磅！微软开源最强小模型Phi-4，超GPT-4o、可商用
去年12月12日，微软首次展示了phi-4，参数只有140亿性能却极强，在GPQA研究生水平、MATH数学基准测试中，超过了OpenAI的GPT-4o，也超过了同类顶级开源 ...
来自MSN14 天
OpenAI为研究长寿推出GPT-4b，联手清华大牛丁胜搞细胞重编程
梦晨 西风 发自 凹非寺量子位 | 公众号 QbitAI OpenAI要研究人类长寿问题了？？推出新模型GPT-4b micro。 可以设计出将普通细胞转化为干细胞的蛋白质。 据MIT科技评论消息，这是OpenAI首个专注于生物数据的大模型，也是OpenAI首次公开声称其模型可以带来意想不到的科学发现。 研究合作方Retro ...
devdiscourse7 天
When AI meets medicine: Advancing multilingual care with GPT-4
One of GPT-4’s key strengths lies in its ability to extract explicit and straightforward information from medical notes. The ...
Neowin9 个月
Microsoft says GPT-4 Turbo with Vision is generally available on Azure OpenAI Service0 0
Today, the company announced that GPT-4 Turbo with Vision is now generally ... such as Optical Character Recognition (OCR), object grounding, video prompts, and "Azure OpenAI Service on your ...
TechNode14 天
Xiaohongshu rolls out AI-powered translation feature, likely using GPT-4
13 to handle the sudden surge in user traffic and translation demands sparked by the so-called #TikTokRefugees trend, which began as the Chinese-owned short video app faced a ban on the US ... and the ...