Change what a model knows or how it behaves, persistently — direct weight edits, ROME-style rank-one updates, persistent activation edits via model.edit, loading PEFT/LoRA adapters, and training a LoRA or adapter through a frozen model with nnsight's interleaved backward. Use to install or overwrite a fact, to bake an intervention into a model everyone shares, to fine-tune a small adapter while watching internals, and to evaluate an edit for specificity and generalization rather than just checking that the target prompt changed.
