Pinned Tweet
《忠诚真的很重要》
那些拒绝你的事物强迫它 只会遭到反噬
那些没有拒绝你的 为什么你要去嫌弃甚至厌恶呢?
那些迎合你的 你甚至不曾多看一眼?
我想我应该反思一下自己 去接受那不拒绝和迎合自己的事物了
那只会拒绝你的 应该被遭到唾弃 甚至应该被厌恶
人不应该原谅真正伤害过你的事
2
11
2,393
Astra 6.2建模的时候我发现他会给图片打方格进行建模
我在Fable5.2也看到这么做了 但是出来的效果明显是Astra6.2更好
GPT生图+Astra6.2根据图片建模我感觉已经可以进入一定的生产了
4
14
1,895
谷谷歌歌干死达里奥!
Gemini 4 Krypton Leaks : RSI is really working
- Gemini 4 Krypton is available on Arena under the codename "gemini-4-carbon".
- Krypton is the best model Google is working on to compete with Fable and Astra level models.
- A new Krypton checkpoint is here, and its already one of the best models currently available.
- Its SVG output is impressive and comparable, and in some places even better than Fable 5.2.
- Its 3D design taste is quite impressive. It handles small details really well, and of course, its design taste is one of the best.
- It takes time to build as its a hardworking model. Its also the best model I have found for research, even better than Fable.
Its mostly coming next month, but Google is shipping very fast, so it can by the end of this month also. Google flexing their RSI.
1
4
346
把老黄赶跑了无法无天了?
The NVIDIA Board of Directors has established the National Security and Export Control Oversight Committee. Read the release: nvda.ws/4eErJ5F
1
2
225
太安全了 安全到我的订阅快被取消了
We’re working with @ApolloResearch, @METR_Evals, @TransluceAI, and @GoodfireAI on embedded assessments: giving outside experts deep access to our internal models and systems to investigate alignment and test our safeguards.
These assessments build on our ongoing alignment and safety research, bringing independent scrutiny to how we identify and manage risks in our systems.
openai.com/index/our-first-p…
5
13
844






























