LLM의 Next-token Prediction 구조와 인간 의식의 인과적 역전 분석
Words Are a Byproduct of Consciousness. For LLMs, It's Backwards
Words Are a Byproduct of Consciousness. For LLMs, It's Backwards
Beyond ChatGPT: Understanding the Core Building Blocks of Generative AI
DiScoFormer: One transformer for density and score, across distributions
How Modern Transformer Blocks Work — From RMSNorm to MoE
Self-Attention: The Brilliant Idea That Made Large Language Models Possible
Transformer 한계 극복을 위한 SSM·JEPA 기반 지속 학습 아키텍처로의 전환
Building a Financial Named Entity Recognition Pipeline for Enterprise AI
How Claude AI Actually Works: The Technical Story Behind the Scenes
Transformers From Scratch: Assembling the Block Behind GPT
When Software Started Writing Software: A Developer’s History of AI
The AI Conundrum: We are living in highly subsidized, interesting times
Three Ideas Made Modern AI Possible. None of Them Are Magic.
Three Ideas Made Modern AI Possible. None of Them Are Magic.
Section 1.1 — Comparing AI Types and Techniques Used in Cybersecurity
How Self-Attention Works — QKV, Softmax, and Matrix Computation
박사급 RS 오퍼 싹쓸이를 위한 ML 시스템 구현 및 인터뷰 최적화 전략
GateGPT: 56k tokens per second Transformer (KV cache) on FPGA at 80 MHz
Intro to Gen AI for Python Beginners: Stop Just ChatGPT‑ing and Start Using ChatGPT
KV Cache in LLMs: The Optimization That Makes Modern AI Models Feel Fast
Understanding Attention in Transformers — Intuition Before Equations