Strategies API

Machine unlearning strategies modify a trained model to remove the influence of specific data points while preserving model utility.

Base Strategy

Gradient Methods

Parameter Methods

Data Methods

VLM-Specific

LLM-Specific

Diffusion-Specific

Ensemble

Strategy Registry

All strategies are registered via the global registry and can be accessed by name:

from erasus.core.registry import strategy_registry

# List all registered strategies
print(strategy_registry.list())

# Get a specific strategy class
StrategyClass = strategy_registry.get("gradient_ascent")