🧠 Transformer Architecture - inarcissuss

Discussed on Substack

🤖人工智能 IT之家·

DeepMind CEO 哈萨比斯回应“谷歌 AI 人才流失”：数据、硬件、算力仍是招聘王牌

Less-relevant results

⚡Transformers 3 Quarks Daily·

AI Collapses on a Classic Psychology Test. What It Reveals Could Stall Human-Level AI

Covers Attention is all you need (2017)

🤖人工智能 wukaipeng.com·

注意力革命: 微信

⚡Transformers Breno·

The self-attention first minimum code

🧠LLM Research medium.com

Positional Encoding: From Sinusoidal to RoPE to ALiBi

⚡Transformers colobu.com·

LLM 究竟是如何工作的？

🤖AI SuperTechFans·

2026 06 24 HackerNews

Covers 8 stories including Canada is looking to build up to 10 new nuclear reactors over the next 15 years

🤖人工智能 weixin.sogou.com·

视频播客怎么就成香饽饽了？

🧪AI Labs mittrchina.com·

13人团队叫板Anthropic：我们造了一个更快更便宜的大模型

🔬Deep Learning Neuroscience News·

Human Memory Limits Make AI Better at Grammar

🤖人工智能 cnbeta.com.tw·

Facebook将Creator Studio改造成AI伴侣应用

🧠LLM Research IEEE Spectrum

IEEE Rolls Out Large Language Models Virtual Training Course

Covers 5 stories including How to Compress DICOM (.dcm) Images from 1.4 MB to KB Using Python?

Covered by contextmaestro.com

⚡Transformers SingularityHub·

AI Collapses on a Classic Psychology Test. What It Reveals Could Stall Human-Level AI.

Covers Attention is all you need (2017)

Covered by 3 Quarks Daily

📄AI Papers arXiv·

RoFormer: Enhanced Transformer with Rotary Position Embedding

Covered by 13 sources including pathtostaff.com, DEV Community

🤖AI medium.com

Transformer Model: The Brain Behind Modern AI

⚡Transformers Nature·

Towards million-token context windows: a topology-preserving framework for adaptive transformer sparsification

🤖ai 应用 guokr.com·

10 分钟就能快速恢复精力，周一上班的你不知道就亏大了

🤖人工智能 RFI - 法国国际广播电台·

Communicability-Inspired Positional Encoding (CIPE)

Tree Transformers

DeepMind CEO 哈萨比斯回应“谷歌 AI 人才流失”：数据、硬件、算力仍是招聘王牌

AI Collapses on a Classic Psychology Test. What It Reveals Could Stall Human-Level AI

注意力革命: 微信

The self-attention first minimum code

Positional Encoding: From Sinusoidal to RoPE to ALiBi

LLM 究竟是如何工作的？

2026 06 24 HackerNews

视频播客怎么就成香饽饽了？

13人团队叫板Anthropic：我们造了一个更快更便宜的大模型

Human Memory Limits Make AI Better at Grammar

Facebook将Creator Studio改造成AI伴侣应用

IEEE Rolls Out Large Language Models Virtual Training Course

AI Collapses on a Classic Psychology Test. What It Reveals Could Stall Human-Level AI.

RoFormer: Enhanced Transformer with Rotary Position Embedding

Transformer Model: The Brain Behind Modern AI

Towards million-token context windows: a topology-preserving framework for adaptive transformer sparsification

10 分钟就能快速恢复精力，周一上班的你不知道就亏大了

特朗普坚称是梅洛尼央求与他合影