There is a one sentence precise definition for a graph theoretical concept that I would like to copy more or less 1:1 into my paper. The only thing I change is some naming as "points" to "nodes" to fit it to the rest of the paper. Is this assumed plagiarism? The definition is short and precise. Changing words or replacing them would prolong the definition unnecessary. If taking the definition as it is would requiring me to reference it I would rather rewrite it and inflate it with words. On the other hand wikipedia hasn't referenced where they have taken the definition from and they are not the first to use it. (Can say so for sure, because it has been around long before.)
Plagiarism is copying text from another source without (proper) attribution. As a quotation, you can put text from outside sources into you paper verbatim if is marked as such (see example).
"A quotation is the repetition of a sentence, phrase, or passage from speech or text that someone has said or written." Wikipedia,2021
If you slightly adjust the text, leave out the quotation marks but still reference the source:
Repeating parts of written or spoken information is considered a quotation Wikipedia,2021.
Same goes for definitions.
Generally speaking, what you find in wikipedia is "Common Knowledge". Wikipedia isn't intended to contain original research or even creative ideas. It presents facts that are known and tries, in many cases, to show where the ideas came from.
What you read there is technically copyrighted, but with a broad license for reuse: https://en.wikipedia.org/wiki/Wikipedia:Copyrights. It is, in fact, intended for reuse.
There is actually a prohibition against posting original research there (https://en.wikipedia.org/wiki/Wikipedia:No_original_research) and, when it is found, it is first marked and then removed.
Therefore, the ideas you get in wikipedia are free to use and plagiarism isn't an issue. However, if you quote from wikipedia you need to cite it. This is actually the same as quoting from any copyrighted work. You quote to indicate that the words came from elsewhere and you cite to make the source clear.
Citation, even of wikipedia, is a benefit to the reader who seeks, in most cases, not only your words, but context in which your words are developed. Wikipedia in general, especially for scientific articles, is good about providing such context.
So, I suggest that you cite wikipedia (judiciously) just as a service to readers, though a search of the scientific terms you use will probably turn up related wikipedia articles as top level hits.
Note that plagiarism is about ideas, not "text". I can plagiarize using literally none of the words in the original. It is misattribution of ideas. Copyright is about the specific expression of those ideas. It is a different concept. If I steal words from a copyrighted source, then I might break copyright, but maybe not plagiarize (common knowledge).
I'll also note that some people are confused about the definition of plagiarism and the other things that govern IP.
Unfortunately, in the context of a classroom, when an instructor sets the rules, things can get confusing, as what is called plagiarism there may not actually be that. It is used as a catch-all for lots of things that involve improper (in the class context) copying. But, I can't "plagiarize" by copying a bubble sort program, since bubble sort is common knowledge, though maybe not to students. I can certainly break the professor's rules, of course.
So, some people get the idea that plagiarism is about the words. It isn't. It is about the ideas behind the words.