近日，上海人工智能实验室旗下司南评测体系OpenCompass发布了令人瞩目的首个AI高考全卷评测结果。此次评测选择了7个顶尖大模型进行高考“语数外”全卷能力测试，满分为420分。
在激烈的竞争中，阿里通义千问2-72B脱颖而出，以303分的高分位居榜首。此外，OpenAI的GPT-4o以296分获得第二名，上海人工智能实验室的书生·浦语2.0排名第三，三个大模型的得分率均超过70%，显示出强大的实力。
值得注意的是，此次高考测试并非所有AI模型都表现优异。来自法国的大模型初创公司Mistral排名末尾，反映出人工智能领域内的竞争依然激烈，技术差距尚未完全弥平。
此次AI高考全卷评测结果不仅展示了人工智能技术的发展水平，也为行业内部提供了宝贵的参考。随着技术的不断进步和应用的不断拓展，人工智能在各个领域发挥着越来越重要的作用。未来，人们对于人工智能的期待和要求也将越来越高。
此次评测结果反映了人工智能领域的前沿动态和发展趋势，让我们期待更多优秀的大模型在未来展现出更加出色的表现。同时，人工智能技术的不断发展和应用将为人类社会带来更多的便利和创新。
英语如下：
News Title: “AI Large Model Examination First Performance: Alibaba’s Tongyi Top Question Takes First Place, GPT-4o and Shusheng Pu Yu Follow Close Behind”
Keywords: 1. AI Examination Evaluation
News Content:
Recently, the OpenCompass, a sub-brand of Shanghai Artificial Intelligence Lab, has released the results of the first AI large model examination evaluation, known as the “AI College Entrance Examination.” This evaluation tested seven top AI models on their abilities across subjects akin to “Chinese, Math, and Foreign Languages” in a high school exam, with a maximum score of 420 points.
In fierce competition, Alibaba’s Tongyi Top Question 2-72B stood out, scoring a high mark of 303 and taking the top spot. Additionally, OpenAI’s GPT-4o secured second place with a score of 296, followed closely by Shanghai Artificial Intelligence Lab’s Shusheng Pu Yu 2.0 in third place. The top three models achieved a score rate exceeding 70%, demonstrating their remarkable capabilities.
It’s noteworthy that not all AI models performed well in this exam. The French AI startup Mistral ranked at the bottom, highlighting the fierce competition within the AI field and indicating that the technological gap has not yet been fully closed.
The results of this AI College Entrance Examination evaluation not only showcase the level of AI technology development but also provide valuable insights for the industry. As technology continues to advance and applications expand, AI is playing an increasingly important role in various fields. People’s expectations and demands for AI will also continue to grow in the future.
This evaluation reflects the frontiers and trends in the AI field, leaving us to expect more outstanding performances from excellent large models in the future. The continuous development and application of AI technology will bring more convenience and innovation to human society.
【来源】https://ai-bot.cn/go/?url=aHR0cHM6Ly93d3cueWljYWkuY29tL25ld3MvMTAyMTU2ODg5Lmh0bWw%3D
Views: 0