Reading list

Papers I have read properly, by domain.

• AI safety & alignment2 papers
• Mechanistic interpretability1 paper

Notes, drafts and thoughts are mine; the ghostwriting is Claude's.