How can I filter nested array object fields when querying Dynamodb?

I have a dynamodb item which looks like:

{
 "taxAmounts": [
  {
   "amount": {
    "currency": "AUD",
    "value": 209
   },
   "name": "GST"
  }
 ]
}

I am looking for a filter to filter on the name field of taxAmounts. I have tired:

--filter-expression "contains (taxAmounts.#name, :gst)"
--expression-attribute-values '{":gst":{"M": { "name": {"S":  "GST"}}}}'
-expression-attribute-names '{"#name": "name"}'

but it doesn't return any item. What is the right way to filter them?
but

taxAmounts is a list, so you need [n] to select the n-th element. "GST" is a string, so no need to bother about the Map data type.

Try this:

--filter-expression "contains (taxAmounts[0].#name, :gst)"
--expression-attribute-values '{ ":gst": { "S": "GST" }'
--expression-attribute-names '{"#name": "name"}'

EDIT

Editing based on your and Hunterhacker's comment. If you want to search the entire list instead of only the n-th element, you may consider re-structuring your list into a map. Example:

{
  "taxAmounts": {
    "GST" : {
      "amount": {
        "currency": "AUD",
        "value": 209
      }
    },
    "TAX2": {...},
    ...
  }
}

You can then search using the attribute_exists function.

FilterExpression = 'attribute_exists(#0.#1)',
ExpressionAttributeNames = {
    "#0": "taxAmounts",
    "#1": "GST"
}

Speaking of re-structuring, scan is an expensive operation in the first place. Filter does not help to reduce the scanned items, as it only filters the result after the scan and before returning the response. Consider having a second table, say taxAmounts, having GST, TAX2, etc. as the Sort Key.