This page is for the honors seminar associated with COMPSCI 446. Please see that page for details about the associated course.
NOTE: If you are hoping to take the honors colloquium part of 446 -- aka H446 -- you must fill out the override form available on the college's override page ("override froms for COMPSCI, INFO, and CICS courses"). Note that the Fall 2025 form will not be available until April 25, 2025.
In this one-credit honors colloquium for COMPSCI 446, students will explore and discuss topics from the 446 curriculum in greater detail, with an intended focus on contemporary issues related to search engines -- for example, large language models, fairness, and/or explainability. Students may also collaboratively design programming projects that build on the programming project from 446.
All students will produce a final report and may implement the expanded programming project to replace part of that report.
The seminar is under development, but will depend upon the course textbook (free) and also upon several articles that will be selected to be freely available online, though some may require that you download them from on campus.
The honors seminar is tentatively slated to meet Thursdays from 2:30-3:20pm in room 140 of the CS building (not the new one). The actual time may change upon discussion with students registered in the class.
Your grade in this class will be based upon completing the readings as demonstrated by class participation in discussions (20%) and your final report (80%). The seminar cannot be taken pass/fail.
This class is open to Commonwealth Honors College junior and senior Computer Science majors only. Students must be concurrently enrolled in COMPSCI 446 or have taken it previously and received a B or better.
If you wish to take this seminar but do not satisfy those requirements (per SPIRE), you can fill out the override form and provide information to request consideration anyway.
You may discuss the ideas behind assignments with others. You may ask others for help understanding class and search engine concepts. You may study with friends. However...
The work that you do and submit must be your own. It may not be copied from the web, from another student in the class, or from anyone else. If you use any ideas that are not your own and that are not clearly well-known, you must cite the source of the idea, even if it is another person. That is true whether you insert a quotation into your writing or just use the ideas you found. If in doubt, cite. If really in doubt, cite and also ask. Failure to cite in any of these cases is plagiarism and grounds for academic discipline.
LLMs and ChatGPT and friends. What is the implication for using large language models and so on? You have to cite your sources, so you need to cite the LLM/GPT system you used and how the resulting answer is appropriate (as with any source), including the prompt you used to find the information.
Although you may find them helpful to pull together some ideas or get you started, you are strongly discouraged to use LLMs/GPTs to write any of your actual final report. Why? It is already well known that those models hallucinate very reasonable sounding material, so you could end up with garbage. And a bad grade. It is also well known that they were trained on existing work and occasionally just spit it out verbatim, so you could end up "accidentally" plagiarizing. And, you recall, plagiarism is grounds for academic discipline, even if it was LLM/GPT that plagiarized "for" you without your knowledge.
Using LLMs to support any code you write for this seminar is acceptable, but (1) it should be for small parts of the assignment such as tossing in a sort algorithm, and (2) you must indicate in your code where you used an LLM and how you confirmed that the resulting code works.