LogoThread Easy
  • 探索
  • 撰写 Thread
LogoThread Easy

您的一体化 Twitter 线程助手

© 2025 Thread Easy All Rights Reserved.

探索

最新在前,按卡片方式浏览线程

开启时会模糊预览图,关闭后正常显示

RT @sainingxie: most of people didn’t know this we had been using TPUs at *Facebook* as far back as 2020. Kaiming led the initial developme…

RT @sainingxie: most of people didn’t know this we had been using TPUs at *Facebook* as far back as 2020. Kaiming led the initial developme…

🇦🇺 Co-founder: @AnswerDotAI & @FastDotAI ; Prev: professor @ UQ; Stanford fellow; @kaggle president; @fastmail/@enlitic/etc founder https://t.co/16UBFTX7mo

avatar for Jeremy Howard
Jeremy Howard
Thu Nov 27 18:11:56
Man, we need much longer contexts… and much better refinement against overthinking. We probably will have to inference a lot of >1M chains before cutting it down where applicable. 
DSA in theory could support it without prohibitive compute costs

Man, we need much longer contexts… and much better refinement against overthinking. We probably will have to inference a lot of >1M chains before cutting it down where applicable. DSA in theory could support it without prohibitive compute costs

We're in a race. It's not USA vs China but humans and AGIs vs ape power centralization. @deepseek_ai stan #1, 2023–Deep Time «C’est la guerre.» ®1

avatar for Teortaxes▶️ (DeepSeek 推特🐋铁粉 2023 – ∞)
Teortaxes▶️ (DeepSeek 推特🐋铁粉 2023 – ∞)
Thu Nov 27 18:10:24
Man, we need much longer contexts… and much better refinement against overthinking. We probably will have to inference a lot of >1M chains before cutting it down where applicable. 
DSA in theory could support it without prohibitive compute costs

Man, we need much longer contexts… and much better refinement against overthinking. We probably will have to inference a lot of >1M chains before cutting it down where applicable. DSA in theory could support it without prohibitive compute costs

We're in a race. It's not USA vs China but humans and AGIs vs ape power centralization. @deepseek_ai stan #1, 2023–Deep Time «C’est la guerre.» ®1

avatar for Teortaxes▶️ (DeepSeek 推特🐋铁粉 2023 – ∞)
Teortaxes▶️ (DeepSeek 推特🐋铁粉 2023 – ∞)
Thu Nov 27 18:10:24
In case you wondered what «DeepSeek» and Longtermism» are about, it's not that they're just spamming random English tokens.

In case you wondered what «DeepSeek» and Longtermism» are about, it's not that they're just spamming random English tokens.

We're in a race. It's not USA vs China but humans and AGIs vs ape power centralization. @deepseek_ai stan #1, 2023–Deep Time «C’est la guerre.» ®1

avatar for Teortaxes▶️ (DeepSeek 推特🐋铁粉 2023 – ∞)
Teortaxes▶️ (DeepSeek 推特🐋铁粉 2023 – ∞)
Thu Nov 27 18:00:38
GRPO without updates is enough to fight Google for IMO Gold
Whales are not in the algorithmic fetishism paper mill business, they're in the «answer the essential question with long-termism» business.

GRPO without updates is enough to fight Google for IMO Gold Whales are not in the algorithmic fetishism paper mill business, they're in the «answer the essential question with long-termism» business.

In case you wondered what «DeepSeek» and Longtermism» are about, it's not that they're just spamming random English tokens.

avatar for Teortaxes▶️ (DeepSeek 推特🐋铁粉 2023 – ∞)
Teortaxes▶️ (DeepSeek 推特🐋铁粉 2023 – ∞)
Thu Nov 27 17:52:54
Now that my skillset is about 80% automated away, I understand the artist reaction to AI even less. This rules!

Now that my skillset is about 80% automated away, I understand the artist reaction to AI even less. This rules!

Opening portals to VR without headsets at @portalvr_io. Problems soluble, potential to improve invariant.

avatar for gfodor.id
gfodor.id
Thu Nov 27 17:52:16
  • Previous
  • 1
  • More pages
  • 2190
  • 2191
  • 2192
  • More pages
  • 5634
  • Next