Centralised simulations

Parameter-server distributed SGD simulation.

Each synchronous round follows the same pattern:

1. Honest workers compute a gradient on their local data shard.

2. Byzantine workers craft adversarial gradients.

3. The aggregator combines all \(n\) gradients into a single update.

4. The aggregated update is applied via an SGD step.

Base simulation class

• CentralisedSimulation

Available simulations

• HiddenVulnerabilitySimulation
• KrumSimulation