Все публикации

0:05:27

Comparing 5 Prompts To Easily Understand AI Research Papers - My micro-research

0:08:24

Upgraded Claude 3.5 Sonnet Can Use Your Computer, 3.5 Haiku Better Than Opus

0:18:02

They discovered 6 reasoning patterns of OpenAI O1

0:10:43

Nvidia just INVENTED a 15x faster Transformer - nGPT

0:19:54

Apple just DEBUNKED LLM reasoning (OpenAI O1, Llama3, OpenAI 4O, Phi...)

0:18:47

Pixtral 12b just broke the ankles of other multimodal models - Paper Review

0:24:41

New way to convert any model into Mixture of Experts

0:17:54

New `Differential Transformer` is 35% faster

0:19:02

Meta AI just tried to recreate OpenAI o1 - Self improving thinking AI

0:30:38

2x Faster Inference - SageAttention: 8-bit Attention For Plug-and-Play Inference Acceleration

0:15:07

15 min to make best possible x.com clone - bolt.new vs claude.ai

0:15:34

Code Podcast Generator Micro SaaS From Scratch - NextJS, Typescript, Groq, OpenAI API, Tailwind CSS

0:25:35

bolt.new + claude.ai = Micro SaaS in 20min - LIVE Coding - NextJS, Typescript, Tailwind Replicate AI

0:28:37

Code Micro SaaS From Scratch - AI Story PDF Generator - NextJS,Typescript, Groq, Replicate AI

6:06:14

Learning 200 Chinese Sentences (Listening & Speaking)

11:42:43

Study Chinese With Me For 12 Hours

0:17:39

Rotational Positional Embeddings & Multihead Attention - Road To AI Researcher Day 7

0:05:17

Tensor Broadcasting Explained in Google Colab

0:10:33

Chain-of-thought only helps with these prompts - Road To AI Scienst: Day 5

0:11:06

Paper Review: Why LLMs Can't Self Correct - My Road To AI Scientist: Days 3-4

0:19:07

Neural Network From Scratch - Road To AI Research Scientist: Day 2 - MNIST NumPy

0:12:05

Training Language Models To Self-Correct via Reinforcement Learning - Road To AI Scientist: Day 1

0:11:05

Rotary Positional Embeddings & Rotation Matrix + Python LLM code

0:02:06

Generates 1500 words per second with Llama3