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

The science has spoken: This is the ideal breeding environment for Homo Sapiens. You may not like it, but this is what peak performance looks like.

The science has spoken: This is the ideal breeding environment for Homo Sapiens. You may not like it, but this is what peak performance looks like.

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 13 02:05:29
Semi-Urgent: Looking for a contracting / consulting gig Area: Full-stack / frontend / security info [at] janemanchunwong [dot] com

Semi-Urgent: Looking for a contracting / consulting gig Area: Full-stack / frontend / security info [at] janemanchunwong [dot] com

“The woman scooping Silicon Valley” — BBC・hacker turned builder + blogger・ex: Threads, Instagram, startups, etc

avatar for Jane Manchun Wong
Jane Manchun Wong
Thu Nov 13 02:00:35
Google DeepMind在Nature上发布了其最新研究，【教AI像人类一样看世界】 让AI不是“只会认图”的机器，而是开始“理解图”的智能体 当前视觉AI应用虽广泛，但其“视觉理解”方式与人类有系统性差异，AI无法像人类一样理解“汽车和飞机都是大型金属交通工具”这种层次概念 研究者通过“找不同”任务比较了人类和AI的视觉理解方式，发现AI更容易被颜色、纹理等表面特征干扰，人类更关注语义层级 于是他们用了三步教AI像人一样思考，将AI的视觉表征调整得更接近人类的认知结构 第一步，基于SigLIP-SO400M，冻结主干网络，在THINGS数据集上训练一个小型适配器，得到“教师模型” 第二步，用教师模型生成一个大规模合成数据集AligNet，模拟人类的视觉判断 第三步，用AligNet训练“学生模型”，使其内部表征更符合人类的语义层次结构 对齐后的模型在模拟人类视觉判断、展现类似人类的“犹豫”不确定性，以及面对新任务时，表现的更稳更准 #VLM

Google DeepMind在Nature上发布了其最新研究，【教AI像人类一样看世界】 让AI不是“只会认图”的机器，而是开始“理解图”的智能体 当前视觉AI应用虽广泛，但其“视觉理解”方式与人类有系统性差异，AI无法像人类一样理解“汽车和飞机都是大型金属交通工具”这种层次概念 研究者通过“找不同”任务比较了人类和AI的视觉理解方式，发现AI更容易被颜色、纹理等表面特征干扰，人类更关注语义层级 于是他们用了三步教AI像人一样思考，将AI的视觉表征调整得更接近人类的认知结构 第一步，基于SigLIP-SO400M，冻结主干网络，在THINGS数据集上训练一个小型适配器，得到“教师模型” 第二步，用教师模型生成一个大规模合成数据集AligNet，模拟人类的视觉判断 第三步，用AligNet训练“学生模型”，使其内部表征更符合人类的语义层次结构 对齐后的模型在模拟人类视觉判断、展现类似人类的“犹豫”不确定性，以及面对新任务时，表现的更稳更准 #VLM

博客：https://t.co/8UDFdmMO4s

avatar for AIGCLINK
AIGCLINK
Thu Nov 13 01:59:22
Healthy dynamical complex systems don’t settle into a stable equilibrium; they surf a regulated critical regime, and dynamic pathologies arise when they’re not controlled back into that regime. I would argue that this statement also holds true for most dynamic and complex system including large language models.

Healthy dynamical complex systems don’t settle into a stable equilibrium; they surf a regulated critical regime, and dynamic pathologies arise when they’re not controlled back into that regime. I would argue that this statement also holds true for most dynamic and complex system including large language models.

Exploring science, cognition, AI, robotics, engineering, dynamical complex systems, more importantly life and beyond.

avatar for Faruk “Frank” Guney
Faruk “Frank” Guney
Thu Nov 13 01:56:32
Primeagen 这个视频给我看乐了 竟然真有人在 express 这种开源项目上 直接点击 README 上面的改动按钮，写上自己的名字，然后提交PR…. 更可怕的是，这是一个 git 和 GitHub 的教学视频

Primeagen 这个视频给我看乐了 竟然真有人在 express 这种开源项目上 直接点击 README 上面的改动按钮，写上自己的名字，然后提交PR…. 更可怕的是，这是一个 git 和 GitHub 的教学视频

https://t.co/r6lgFwii8j

avatar for 艾略特
艾略特
Thu Nov 13 01:55:59
RT @craigsdennis: Yo it's @threepointone 's bday🎈 Tell him he looks nice!

RT @craigsdennis: Yo it's @threepointone 's bday🎈 Tell him he looks nice!

Have questions, or building something cool with Cloudflare's Developer products? We're here to help. For help with your account please try @CloudflareHelp

avatar for Cloudflare Developers
Cloudflare Developers
Thu Nov 13 01:54:03
  • Previous
  • 1
  • More pages
  • 24
  • 25
  • 26
  • More pages
  • 2127
  • Next