3903 位用户此时在线
- 本站自动实时分享网络热点
- 24小时实时更新
- 所有言论不代表本站态度
- 欢迎对信息踊跃评论评分
- 评分越高，信息越新，排列越靠前
Grok 4 Heavy (left) V.s. Gemini 2.5 Pro (right)
Create a Turing-complete Scheme interpreter in C that supports lexical scoping, closures, continuations, and proper tail-call for tail recursion without stack growth.
Grok4 won. It wrote superior code.
Grok4 Heavy: 903 Lines of C
btc
(
twitter.com
)
Grok 4 Heavy (left) V.s. Gemini 2.5 Pro (right)
Create a Turing-complete Scheme interpreter in C that supports lexical scoping, closures, continuations, and proper tail-call for tail recursion without stack growth.
Grok4 won. It wrote superior code.
Grok4 Heavy: 903 Lines of C code.
Gemini 2.5 Pro: 891 Lines of C code.
Both compiled!
The code from Grok 4 Heavy worked flawlessly.
The code from Gemini 2.5 Pro did not work even after multiple prompts.
Grok 4 Heavy: ~10 minutes single prompt.
Gemini 2.5 Pro: ~2-3 minutes per prompt after about 10 prompts I gave up.
Full Prompt👇
点击图片查看原图
We spotted a couple of issues with Grok 4 recently that we immediately investigated & mitigated.
One was that if you ask it "What is your surname?" it doesn't have one so it searches the internet leading to undesirable results, such as when its searches picked up a viral meme
btc
(
twitter.com)
Grok 4 Fast is the most fastest smartest AI
It fires up instantly and performs a deep web search going through 100+ web sources and live 𝕏 posts when needed and does it all at incredible speed
The one model that works for everything and it's an absolute beast in performance
btc
(
twitter.com)