Keep on to blur preview images; turn off to show them clearly
RT @ch272h: 🔗 Why knots? Knots are simple to see but deep to reason about. ✔ Verifiable outcomes ✔ Structured complexity (crossing number…
making models learn • eXperiments lab • full of myself, but still hungry
God damn it... ended up building an email client 🤣 What can of warms did I just opeeeeeeeen... habibies save me
CEO @Plutio_app - on a mission to kill busywork with the world’s first Super Work AI. Ex–Head of Brand @ClickUp.
RT @2112Power: Full, unedited @SecWar speech at @HadrianInc. SW Hegseth gets it - we will out manufacture and out scale our adversaries. I…
Co-Founder, American Dynamism. General Partner @a16z. Catholic. Mother. American. 🇺🇸 🚀💪
RT @derangineer: awesome vibes at the modal x event, thanks @modal for awesome NeurIPS events throughout the week. shoutout to @charles_irl…
gpu enjoyer at @modal. he/him. ex @full_stack_dl, @weights_biases (acq. @CoreWeave), phd Berkeley @Redwood_Neuro. try https://t.co/SYWVMCazZ3
mixture of experts is such an ugly idea, i hope it dies and gets replaced by something more fundamental
Conditional computation is a wise goal. It is MoEs in particular that are ugly execution.
Agili 的 AIGC 周刊（Y25W48） 这是一个激动人心的一周。开源大模型继续挑战闭源模型的霸主地位,DeepSeek-V3.2 在多个国际竞赛中达到金牌水平,Mistral 3 系列全面开源。与此同时,Anthropic 以 10 亿美元年化收入的成绩收购 Bun,标志着 AI 编码工具进入新纪元。 https://t.co/BnjrGENqff
前端切图仔🧑🏻💻，后端三脚猫🤷🏻♂️，运维挖坑人🤦🏻♂️，AI 门外汉🧐。 开源: ⚡ https://t.co/wk0K4h9gd5 玩具: 📮 https://t.co/VTi1aFKE2y 🧰 https://t.co/cRxrFndEFH 我在使用的工具： 🛠：https://t.co/47QLYWyxJA