Svelte Hacker News logo
  • top
  • new
  • show
  • ask
  • jobs
  • about

Flash Interpretability: Decoding Specialised Feature Neurons in LLM

arxiv.org

1 points by delduca 8 hours ago