The Remove function will remove matching values from a multi-value token.

In this example, we have one token with a series of values saying One, Two, or Three.

Then we run the Remove function to remove the One values.

The end result is just the Two and Three values remaining.

For removing CR LF, I usually use a Pattern Matching activity to gather everything except the \r and \n elements.

If you set it to "All matches (combined with no space)" you'll get a single token like the one shown in the screenshot.

But if you set it to "All matches (as a multi-value token)" you'll get a multi-value token, which you can then use the Index function of the token editor to do different separators (like space or semi-colon).

A little annoying to use two activities for it, but it gets the job done, and I've effectively replaced the line breaks with semi-colons in this example.