Baidu, China’s leading search engine provider, has announced that its latest AI model, Ernie 3.5, has surpassed ChatGPT on multiple key metrics.
In a statement, Baidu said that Ernie 3.5 had outperformed ChatGPT in comprehensive ability scores and in several Chinese capabilities. The company cited a test that was conducted by the state newspaper China Science Daily using datasets including AGIEval and C-Eval, two benchmarks used to evaluate the performance of AI models.
Baidu’s Ernie 3.5 model is built on top of the company’s previous Ernie 3.0 model, which was already one of the most powerful AI models in the world. The new model has been trained on a massive dataset of text and code, and it has been fine-tuned to perform a variety of tasks, including natural language understanding, machine translation, and question answering.
Baidu says that Ernie 3.5 comes with better training and inference efficiency, which positions it for faster and cheaper iterations in the future. The company also says that the new model will support external “plugins,” which are additional applications that will allow Baidu’s AI to work on more specific scenarios, such as summarizing long text and generating more precise answers.
The sources for this piece include an article in Reuters.