Înapoi la articole →

Cel mai recent articol

Ultima verificare

Urmăritori 0 urmăritori
Articole: 3
  1. Update March 12, 2026: Larger rewrite for improved clarity TLDR The map of how the activations of one ‘source’ layer in an LLM impact the activations in some later ‘target’ layer can provide vectors for steering LLM behavior.

  2. Recent discussion about the tension between mechanistic understanding and applied safety outcomes got me interested in steering vectors.

  3. Interpreting Machine Learned Models Mechanistic interpretability has been gaining prominence as the technical approach to AI safety and alignment.