返回探索

在LLM前加入Jev反射层筛选信息

作者分享在LLM前插入一张Jev(反射)卡片后,Jev会整理worker报告并只向Commander Astra传递必要信息,使相同交互的token消耗从330k大幅下降。

Maybe it's time to stop feeding everything to the LLM (brain) The result of inserting one "Jev (reflexes)" card in the front stage is insane ・Jev sorts the workers' reports ・Only passes necessary info to Commander Astra With just this, the same exchange goes from 【330k →

在LLM前加入Jev反射层筛选信息 1
· 1216 次赞在 X 打开