Not exactly.  Tokens are not supported for unidentified document names (the dialog for configuring Unidentified Documents gives you a hint about that by not providing a token menu or underlining):


However, if you have the document classification module, there is a workaround: you can use a Document Class with no First Page Identification conditions, placed after all other document classes, to "catch" all pages that would otherwise be unidentified.  That document class can then use global tokens (e.g., Date), tokens from Pre-Classification Processing, and tokens from its Page Processing section. 

Tokens from other document classes won't be available (except via Token Collector/Retriever processes) because the processes in those document classes were not run for the current page.