In Prefect 3, task run cached data isn't stored in a specific database table. Instead, caching is managed through cache keys that are computed based on task inputs, the task code definition, and the flow or task run ID. These cache keys allow Prefect to determine if a task can reuse results from previous executions without actually running the task code again.
For more information on configuring task caching, you can refer to the
Prefect documentation on task caching.