Baidu responded to the dispute over the function of Wen Xin: the ability comes from the cross-modal large model, and the use of data is in line with industry practice.

Thanks to CTOnews.com netizens for the delivery of the clues of Frost Wind Shadow! CTOnews.com March 23 news, Baidu today released a note on "Wen Xin Yi Yan Wen Sheng Picture function related feedback", saying that Wen Xin Yi Yan is a large language model developed by Baidu itself, and the ability of Wen Xin to draw a picture comes from the Wenxin cross-modal model ERNIE-ViLG.

Baidu said that in the large model training, it uses the Internet to disclose data, which is in line with industry practice. "you will also see Baidu's self-research strength from the following iteration of rapid tuning of the ability of Wenshengtu."

In addition, Baidu said that Wenxin word is constantly learning and growing in the process of everyone's use, please give self-research technology and products a little confidence and time, do not spread rumors, but also hope that Wenxin word can bring more joy to everyone.

CTOnews.com has reported that Baidu held a press conference on March 16, officially launched the big language model Wen Xin Yi Yan, and demonstrated the comprehensive ability of Wen Xin Yi Yan in multiple usage scenarios. In terms of multimodal generation, Robin Li, founder, chairman and CEO of Baidu, demonstrated the ability of Wen Xin to generate text, pictures, audio and video at that time.