new deepseek v4 pro is now open weight on hugging face (mit license) "v4" is a bit misleading, previous model was only a preview and this one has way ...

DeepSeek V4 Pro开源发布，MIT许可，训练量远超预览版，接近V4.5。

深挖 →

15XX · List2 天前产品88

谷歌DeepMind推出SL2T：手语实时转英语，Pixel 11可免打字登录

Google DeepMind just shipped a model that lets people sign into a Pixel 11 instead of typing. SL2T starts with ASL-to-English in Gboard and Live Trans...

谷歌DeepMind发布SL2T模型，可将美式手语实时翻译为英语，支持Pixel 11免输入登录。

深挖 →

16XX · List2 天前研究85

AI上下文压缩器严重丢失用户规则，评估套件COMPINT揭示性能退化

A great read if you an AI dev. Current context compactors retain 17% of the standing rules users give them. Session Constraints are instructions like ...

研究发现AI上下文压缩器仅保留17%规则，常使任务效果更差。

深挖 →

17XX · List2 天前模型85

DeepSeek再次移除V4-Pro-0813权重，疑因舆论压力

DeepSeek removed the weights of DeepSeek-V4-Pro-0813 again I guess the backlash got to them

DeepSeek再次下架V4-Pro-0813权重，或因舆论压力所致。

深挖 →

18XX · List2 天前产品85

这真是史上最棒的AI设计广告

This is unironically the best ad for AI design ever made

谷歌设计团队用AI制作庆祝动画，展示AI设计潜力。

深挖 →

19XX · List2 天前产品85

谷歌Pixel 11发布：Rambler语音输入等AI新功能亮相

Can’t wait for everyone to try Rambler — one of my favorite models! As someone who has wrist issues and often voice types, this feature is a game ch...

谷歌发布Pixel 11，主打Gemini智能与Rambler语音输入等新功能。

深挖 →

20XX · List2 天前实践82

我在Figure犯的最大工程错误：腱绳驱动手

The biggest engineering mistake I made at Figure was building a tendon-based hand Our first hand design in 2022 was a tendon hand for our F.01 robot. ...

Figure创始人复盘腱绳驱动手部设计是重大工程失误，反思技术选型教训。

深挖 →

21XX · List2 天前产品78

Sakana Chat大升级：免登录免费，支持代码执行

Sakana Chat just got a big upgrade. No login required, free to use: https://chat.sakana.ai/ Powered by Fugu and Namazu, our Japanese LLM. With newly a...

Sakana Chat升级，免登录免费使用，支持代码执行，可快速生成交互应用。

深挖 →

22XX · List2 天前实践78

与Hamel对谈：开源模型何时才是正确选择

Always a good time when I can sit down and yap with Hamel. Not pictured: this 45 minute edit started as an hour half when we talked about everything f...

Hamel与Lambda探讨开源模型适用场景及部署经验，强调多数团队忽视前沿API之外的选项。

深挖 →

23XX · List2 天前实践75

FWIW, lots of VCs also lie or overstate their background. More commonly they dramatically exaggerate their position in rounds/impact on portcos. We wo...

VC行业普遍存在夸大背景和影响力现象，不应聚焦于特定演示。

深挖 →

24XX · List2 天前产品75

掌控每个音符与细节：强大浏览器DAW上线

Own every note, knob and detail in our powerful browser-based DAW. 🎛️

浏览器DAW上线，掌控每个音符与细节

深挖 →

25XX · List2 天前研究75

别信ARC-AGI宣传：其基准并非独有

don't believe the ARC-AGI propaganda that their benchmarks show some capabilities that no other benchmark captures most math or long-context benchmark...

反驳ARC-AGI基准独特性，指出多数数学和长上下文基准同样体现思维模型转变。

深挖 →

26XX · List2 天前研究72

AI实验室下一步论文方向猜测

«I think the next paper is likely on harnesses and agents, swarms, context-level continuity. maybe autoresearch. Hopefully lifetime learning/continue...

作者推测AI实验室下一步论文方向，可能涉及智能体、群体、上下文连续性及终身学习。

深挖 →

27XX · List2 天前实践72

While I am bias, I am unironically listening to my own talk again as I'm making breakfast. So, even though it is my own talk, I can stand my own voice...

作者自嘲听自己的演讲做早餐，并推荐Lambda关于开源模型选型与部署的讨论。

深挖 →

28XX · List2 天前行业72

i'm so fucking confused, am i violently misunderstanding what ETFs are or is this like a founder bragging about "features released" is there any corre...

质疑Corgi Invest宣称ETF数量将超贝莱德，困惑ETF数量与资产管理规模/收入的关系。

深挖 →

29XX · List2 天前模型72

DeepSeek V4令人失望：发布延迟、规模缩水、涨价伤用户

DeepSeek has been very disappointing this year V4 came much later than expected and was smaller than expected. The price hikes also hurt. They are cur...

DeepSeek V4发布不及预期，价格上调，国内排名下滑至四五位。

深挖 →

30XX · List2 天前行业65

特斯拉眼中的印度交通：狂野又令人唏嘘

Wild and sad!

特斯拉视角下的印度交通视频，展现混乱路况，引发感慨。

深挖 →

31XX · List2 天前会议65

I am very sorry! #ImSorryWomen

AI生成道歉内容引发争议，网友热议。

深挖 →

数据更新于 2 天前 · 共 2257 篇 · 公众号 + 海外 + 论文自建抓取