~/ai-muninn
search
⌘K
❯
esc
Loading…
blog
model anatomy
github
中
~ / blog
/
tag / deepseek
❯
grep -r "#deepseek" ~/blog
2 matches
date
read
title
2026-11-22
15m
[Benchmark] Pi vs dsh vs Hermes on one local model: dsh used 1.97× Pi's input tokens on small tickets, and the gap closed on long ones
#local-llm
#deepseek
#pi
#deepseek-harness
2026-09-16
10m
[LLM Deep Dive] What actually happens inside one transformer layer — four real models, four data flows
#llm
#transformer
#moe
#qwen3
← back to all posts