[ISSUE] Quality Monitor API Dashboard does not apply slicing expressions

[nbhushan-eneco]

Description
I would like to use Lakehouse monitoring to monitor a forecasting model (using the inference log).
While creating the monitor I specify the slicing expressions to be a list of strings.

Reproduction

  timestamp_col: str = "timestamp"
  granularities:list[str] = ["1 day"]
  model_id_col: str = "model_name"
  prediction_col: str = "forecast"
  label_col: str = "target"
  problem_type: MonitorInferenceLogProblemType= MonitorInferenceLogProblemType.PROBLEM_TYPE_REGRESSION
  ...
  slicing_exprs: list[str] | None = ["horizon_category", "model_name"]

While viewing the deployed dashboard I notice that the slicing expressions are not applied, for example in the granularity_selected table, the sql query to display metrics includes..

...
AND isnull(slice_key) AND isnull(slice_value) -- default to "No Slice"
...

Expected behavior
The generated dashboard allows me to slice and display metrics per horizon_category and model_name.

Is it a regression?
N/A

Debug Logs
N/A

Other Information
N/A

Additional context
https://databricks-sdk-py.readthedocs.io/en/stable/workspace/catalog/quality_monitors.html