benchclaw-stage2-simulator-gt-materialization

Ground truth materialization for simulator tasks processes simulator outputs into structured sets using parameters like simulator_id and task_family. This ensures accurate training and evaluation data for AI models.

Materializing ground truth requires specifying simulator_id, simulator_card_skill, and task_family parameters. These inputs identify the simulator context and task execution details for accurate GT record association.

You create a structured ground truth set when a simulator task is executed and you require accurately represented data for training or evaluating AI models. This ensures benchmarking outputs are correctly linked to task execution.

To ensure ground truth records are correctly linked to task execution, use a GT validation process that checks associations based on simulator context and task_family. This guarantees accurate data representation for benchmarking.

Ground truth materialization works with simulator task families identified by required input parameters. You must specify the simulator_id and task_family context to correctly generate and validate the structured GT set for your environment.