Schalk Neethling
Schalk Neethling @scriptquiz ·
I did some research and experimentation. What I learned was truly interesting. Auto-Improving an Agent Skill: Applying @karpathy's Autoresearch Pattern to Semantic HTML schalkneethling.com/posts/auto-imp… #ai #codingwithai
Auto-Improving an Agent Skill: Applying Karpathy's Autoresearch Pattern to Semantic HTML

How I built an eval-judge-improve loop to autonomously refine a semantic HTML agent skill, taking it from 2.46 to 2.89 out of 3.0 across four iterations and what I learned about the limits of...

From schalkneethling.com
19
Eric Risco
Eric Risco @ericrisco_dev ·
Los LLMs pueden tener 4 tipos de alucinaciones al programar. El peor es cuando los errores se acumulan, creando un castillo de naipes frágil. Protégete de estas fallas. #LLM #CodingWithAI #Alucinaciones
24