You manage a Fabric workspace where a data engineering team runs Spark jobs with highly variable data volumes—some nightly batches process a few gigabytes while month-end batches process several terabytes. The team complains that jobs either fail from insufficient resources during peak loads or waste capacity during light loads. You want to configure the workspace Spark settings so that each job automatically adjusts the number of executors within a defined range based on workload demand. Which configuration should you enable?