攻城狮/业余投机/右侧交易/游戏开发/Haskell/Rust/C++/Unity3D/C#/对java有偏见/RL/智商欠费/浅尝辄止故平庸/反乌托邦/竹林中/地狱变/手撕菠萝蜜/胸口碎榴莲/单机推特中/乐视一生黑(乐视已阵亡)/华为一生黑(迟早会阵亡 )/中华跪族/器材党/预防式B台支

上海, 中华人民共和国
Joined June 2012
Include

Only show posts containing:

Exclude

Hide posts containing:

Time range
-
Minimum likes
日常焦虑帝
@gpuhell
2h
Muse is basically a data harvester that goes out of its way to collect your personal information, while offering very little practical value. Straight to the trash.
1
3
96
日常焦虑帝
@gpuhell
Sep 27
Replying to @ScarletKc
OAI 和 A\ 都没被封,都两网站刚上线时注册的老号。
1
5
2,579
日常焦虑帝
@gpuhell
Sep 27
Replying to @jackbilldarren
今天打开 APP 直接注册就好了啊,哪来这么多戏
1
1
85
日常焦虑帝
@gpuhell
Sep 25
你在台湾也这样吗
1
1
38
日常焦虑帝
@gpuhell
Sep 24
Replying to @wquguru
还怪好听的 👍
1
1
108
日常焦虑帝
@gpuhell
Sep 24
my timeline after opus 5.5 release
1
2
132
日常焦虑帝 retweeted
🔥K3.1已确认泄露!! Coming Soon~
37
10
4
231
22,401
日常焦虑帝
@gpuhell
Sep 23
Replying to @0xLogicrw
用意很明显了。。。🤔
1
1
137
日常焦虑帝
@gpuhell
Sep 23
Only Google has truly responded to the call of "Pace the Frontier." 😆
只有谷歌真正响应了ai发展减速的号召。
1
1
4
280
日常焦虑帝
@gpuhell
Sep 22
Replying to @Kimi_Moonshot
工具多解决不了用量抠门的问题
73
日常焦虑帝
@gpuhell
Sep 22
Replying to @mranti
很容易因此收到 cybersecurity abuse 的警告邮件。
1
1,769
日常焦虑帝
@gpuhell
Sep 22
Replying to @kuroyei
宋浩的讲的段子比知识多...
1
601
日常焦虑帝
@gpuhell
Sep 22
Qwen 4 is coming soon. Qwen 4.5 and Qwen 5 are targeting a parameter size of 4–10T.
14
821
日常焦虑帝
@gpuhell
Sep 22
Replying to @AgiWhen
they decided to release the checkpoint 😀 nitter.cf/XiaomiMiMo/status/2102…
Introducing Xiaomi MiMo-V2.6 — Pro & Flash. Frontier intelligence, all the modalities, built in public. 🔹 Two omnimodal models, advancing through scaled reinforcement learning 🔹 Pro performs on par with Claude Opus 5 and GPT-5.6 Sol across most agent benchmarks 🔹 Pro scores 46 on the Artificial Analysis Intelligence Index — the highest among open-source models 🔹 Stronger coding, computer use, 3D reasoning and creative capabilities 🔹 Open model weights, technical report, RL environments and training code Blog:mimo.xiaomi.com/mimo-v2-6
20
日常焦虑帝
@gpuhell
Sep 21
The previously missing data from MiMo’s RL training has now been added.
日常焦虑帝
@gpuhell
Sep 21
MIMO’s RL training has stopped at step 30. We can observe the following: The Pro model achieved a DeepSWE score of 72.57, but this score was reported at step 28; data for steps 29 and 30 are missing. The number of active environments for Pro began increasing at step 23, then dropped rapidly after step 26. The duration of each training step also rose sharply.
5
624
日常焦虑帝
@gpuhell
Sep 21
76
日常焦虑帝
@gpuhell
Sep 21
MIMO’s RL training has stopped at step 30. We can observe the following: The Pro model achieved a DeepSWE score of 72.57, but this score was reported at step 28; data for steps 29 and 30 are missing. The number of active environments for Pro began increasing at step 23, then dropped rapidly after step 26. The duration of each training step also rose sharply.
2
1
5
989
日常焦虑帝
@gpuhell
Sep 21
Replying to @nrehiew_
mimo-v2.6-pro stopped at step 30. but the DeepSWE score 72.57 is from step 28.
4
488
日常焦虑帝 retweeted
Looks like the MiMo RL runs have completed. The pro model went from 58.41 to 72.57 on DeepSWE. For context, the highest score on DeepSWE is 74 by Astra, Gemini 3.8 Flash and Opus 5
27
21
12
641
113,684
日常焦虑帝
@gpuhell
Sep 21
Replying to @ZixuanLi_
此时,智谱公关团队的心情:
1
48
1,833