Can co-authoring a paper as a new grad harm my reputation/career?

I somewhat disagree with the previous answers. I think it is largely improbable that publishing a paper would harm your career, but it could if, for example, it is really poor and someone happens to read it.

Eykanal in a comment to Dave Clarke said that "no one would consider holding you accountable for the content of the paper"; I recall that there is a important trend to insist that all authors of a paper should be accountable for its content.

The fact that you do not want to pursue in the direction of the paper is completely harmless, though. There is really no problem working in different areas, the only thing to be careful about is not to spread oneself efforts too much, but this does not apply here.

So the main issue is (a); here I would say that it can be difficult to judge the quality of a result, especially for an undergraduate, and I would advise to trust your advisor. So it is really, really unlikely that co-authoring this paper could do any kind of harm to your career; in fact it could do more harm to decline this opportunity, since your advisor would probably not understand and you will probably need his or her recommendation.

At the end, I do not disagree that much with other answerers; but I would be less general in my statements.

In short, unless there are ethical concerns, which is unlikely to be the case in computer science, then I'd say no. If you get some paper published as a Bachelor student, this demonstrates your ability to do research, which is what people in charge of admissions are interested in. If the paper gets accepted at a good venue, then this is even better.

At the current stage of your career, your main concern should be getting into a good graduate program. After you write more papers, better ones on the topic you choose, no one will even worry about that first paper.

Yes, participating in a research paper could harm your reputation/career.

But it's very unlikely, as long as you take basic precautions:

• Avoid quack journals, crank journals, and the like
• Avoid ethical breaches (plagiarism, fraudulent data)
• Avoid co-authoring with known cranks

As to your specific issue: "I'm not convinced of the quality of the suggested solution." - if it does indeed solve the problem, then publishing is fine. If it may not, then you need to work things through with your co-author until you agree on whether or not it does solve the problem.