Fast Technology News on June 10. Recently, the China Enterprise Development Research Center of the Xinhua News Agency Research Institute conducted an experience evaluation of mainstream large-scale model products and released the “Artificial Intelligence Large-scale Model Experience Report”.
“The report shows,Baidu Wenxin’s total score ranks first in China, and ranks first in the country in terms of basic ability, IQ, EQ, and work efficiency.
Advertisement
In the part of the IQ test, Baidu Wenxin unexpectedly surpassed ChatGPT3.5with outstanding performance, ranking second only to GPT4.
In terms of comprehensive index evaluation, this evaluation selects 4 dimensions (basic ability, IQ test, EQ test, and work efficiency improvement ability), 36 sub-ability, and a total of 300 questions to test the current mainstream large-scale model products.
At the same time, the Institute also invited relevant experts to form an evaluation team to deeply analyze the semantic understanding, knowledge reserve, and logical capabilities of each product, and finally obtain the comprehensive index evaluation of each manufacturer’s large model.
Advertisement
The comprehensive evaluation results show that:The indicators of the ChatGPT series models are excellent, and the capabilities of the Chat-GPT4.0 version have been improved to a certain extent on the basis of the 3.5 version.
Wenxinyiyan, an artificial intelligence large-scale model developed by Baidu, is an advantageous product among large-scale models independently developed in China.The rest of the large-scale model products also perform well in terms of basic capabilities, but there is still room for improvement to varying degrees in the face of more complex work content or emotional intelligence environments.
[End of this article]If you need to reprint, please be sure to indicate the source: Kuai Technology
Responsible editor: Xpress