yimoEx

yimoEx

none
HomeArchives

【News】First Experience with Grok3

26
AI Translation
This post is translated from Chinese into English through AI.View Original
AI-generated summary
The tests conducted on Gemini2.0Pro and DeepSeek-R1 were primarily for entertainment purposes. The first test showed both models passed, while the second involved a classic comparison. The third test focused on cognitive abilities. Overall, the results were underwhelming, possibly due to being in the Beta stage, with intelligence levels comparable to Gemini-2.0-flash. There is hope for the development of a better model in the future.

Test Source: Arena

Statement: The test is for entertainment purposes only.

First Test:
【Gemini2.0Pro and DeepSeek-R1 both passed the test】

image

Second Test:
Classic Comparison

image

Third Test:
Cognition

image

Summary: It didn't give me an amazing feeling, maybe it's in the Beta stage?
It feels like the intelligence level is around Gemini-2.0-flash
But I still hope they can create a better model.

Loading...
Ownership of this post data is guaranteed by blockchain and smart contracts to the creator alone.