For university course work, I approach this question not from the perspective of where the questions are directed. I approach it instead from the point of when the questions are asked and how they are phrased.
The first stage of a solving a problem is to define the problem. The goal is to have a statement or statements, a picture, and all of the parameters to the problem in one place. The idea here is to be able to "put the book away", where "the book" is the defining source of the problem.
The second stage of solving a problem is to gather information. The goal is to collect values that are as yet undefined but otherwise anticipated to be part of the solution. This is the point where additional insights are added to the definition. Indeed, in some sense, the second step often runs in parallel with the first step.
The third stage of solving a problem is to propose solutions. This is the brainstorming step. The goal is to reach a point where the statement is made "I am convinced that I can solve this problem completely" or, alternatively, "I am convinced that this problem cannot be solved as stated". This step is where assumptions are stated. This is where equations are proposed. This is where guesses are made and leaps of insight are taken. Any proposed approach that does not violate laws or ethics or standards or rules or regulations is permitted. Approaches that are fundamentally incorrect or improper are to be avoided.
Sometimes, a problem statement is in itself so direct that we combine the first three steps. We have no need to reconsider the problem statement, gather any other information, or indeed brainstorm on potential assumptions. The solution to the problem is essentially that we are to do one thing and one thing only.
The final stage is to evaluate an answer. The goal is to take the best proposed approach and complete it. Here is where the calculators are finally allowed to come out. Here is where the format and layout requirements for the report are dusted off and put in to practice.
The first two stages should be done with copious back-and-forth dialog and questions. The only way to solve the actual problem at hand is to be certain that you understand the problem at hand completely before you start to work on it. Whether you ask for help from the course instructor (preferred) or an expert on-line or your peers does not matter. Your goal when you ask questions is to fill in the gaps in your understanding to reach the point where you know that you understand the problem statement completely. The type of questions that you ask at this point are framed accordingly. Basically ... Do I really understand what this problem is saying and what information I have to try to address it?
The third step (proposing solutions) can be done within a team environment, especially when it involves brainstorming or debating the viability of different assumptions. What should not be done at this point is to ask questions about whether a proposed solution is or is not correct. Rather, questions should be framed about whether the proposed solution "makes sense" given the system at hand, the information at hand, and the assumptions made. Again, whether you ask the course instructor (preferred), an expert, or your peers is not relevant. However, understand at this point, the best answer to the question about whether a proposed solution "makes sense" generally comes first and foremost from the course instructor, it generally comes last if at all from your peers in the course.
The last step of problem solving is typically to be done straight up. This means, sit down and do what you are supposed to do to complete the proposed solution. Do not ask questions, just do it. As needed, consult additional references for the proper methods to just do it.
Team Work versus Group Work
I must conclude with a few comments about team work versus group work in a university environment. This is especially with regard to the question about plagiarism.
Team work is the effort taken by students together to do the first three steps in problem solving. Teams get together early, analyze the problem statement and the information given, debate the pros and cons of various approaches, and ask questions of the instructor, experts, and their peers. Then, all team members go home and do the work on their own. Because each student is fully responsible for their own efforts to "just do it", no one has plagiarized anyone in the case of team work.
Group work is the effort taken by students to collect "the right answers" to problems. Groups start with everyone going home to "do" the problem by themselves. Groups end by everyone getting together to compare answers and collect only the best answers to submit. Because all students put answers from the group in as their own effort, everyone has truly plagiarized everyone else in this case.
Team work is to be encouraged and applauded. Group work is to be shunned and duly penalized. I might give an extreme example of three students submitting an answer to the same problem from a course assignment. Each student in a team of three students deserves full credit for their distinctly different answers to the problem. By comparison, each student in a group of three students deserves to have their common grade cut by 1/3 for submitting clearly identical answers to the problem.
The other aspects of plagiarism involve lessons on how to cite sources of information adequately and properly. The lessons are an entire treatise to themselves.