Notices

  • Week 10 Guest Lecture: Bugs Begin, Bugs Begone

    Posted by Zahra Donyavi Sunday 15 November 2026, 10:26:19 AM.

    Dear students,

    In Week 10, we’ll have a special guest lecture by Dr. Hammond Pearce. Please note that this lecture will start at 3:00 pm on Wednesday, instead of our usual 2:00 pm start time.

    Title: Bugs Begin, Bugs Begone: Large Language Models and Cybersecurity

    Abstract

    Software developers—both human and AI-assisted—can introduce cybersecurity vulnerabilities into their code. Emerging "smart" code agents, such as OpenAI's ChatGPT/Codex, GitHub Copilot, and Anthropic's Claude, can unintentionally contribute to these issues. However, AI-enabled tools also hold significant potential in detecting and mitigating security flaws.

    In this research talk, Dr. Pearce will explore the effectiveness of AI-driven solutions for identifying and repairing vulnerabilities. He will discuss the role of prompt engineering in both exacerbating and mitigating security flaws in AI-generated code and examine how the robustness of large language models can be measured for this purpose. His findings demonstrate that while LLMs can aid in cybersecurity bug repair, much work remains to be done in optimizing their reliability.

    About the speaker

    Dr. Hammond Pearce is a Senior Lecturer at UNSW Sydney's School of Computer Science and Engineering. Prior to this, he served as a Research Assistant Professor at NYU's Department of Electrical and Computer Engineering / NYU Center for Cybersecurity and completed a research internship at NASA Ames. His research focuses on cybersecurity, hardware and embedded systems design, and the intersection of AI with industrial informatics—particularly in the evolution of design processes in hardware and firmware development.

    Dr. Pearce is passionate about investigating how AI-driven tools, such as ChatGPT and other large language models, influence hardware and software development as well as social media. His research has earned several accolades, including the 2026 NSW Young Tall Poppy award, the Cybersecurity Award 2023 for the Best Machine Learning and Security Paper, the Efabless AI-Generated Open-Source Silicon Design Challenge, and the Distinguished Paper Award at the IEEE Symposium on Security and Privacy (2022). He has received funding from organizations such as Australia's ARC, CSIRO, Google, and Intel Corporation. Dr. Pearce obtained his Ph.D. in 2020 from the University of Auckland, New Zealand.

    We hope you can join us.

  • COMP9444 – Project Submission Deadline and Guidelines

    Posted by Fatemeh Barazesh Morgani Wednesday 11 November 2026, 11:15:31 AM, last modified Wednesday 11 November 2026, 11:17:00 AM.

    Dear Students,

    The deadline for the COMP9444 Project Submission is:

    Monday, 23 November, 5:00 PM

    Please carefully follow the guidelines below for your submission.

    What to Submit

    Each team must submit the following:

    • Project code (Jupyter Notebook)
    • Project summary report (in .docx or .pdf format)
    • Presentation slides (in .ppt or .pdf format)

    You may compress these files into a single ZIP file before submitting.

    How to Submit

    Please email your submission to your tutor. You can find your tutor's contact information here:
    Tutorial Schedule

    Please:

    • CC all team members (to ensure transparency and consent)
    • CC cs9444@cse.unsw.edu.au
    • Use your group name as the email subject (e.g., Group 23 – Project Submission)

    Due to potential file size limitations, please do not attach large files directly to the email. Instead:

    • Share the files via Google Drive or Microsoft OneDrive.
    • Make sure your tutor has permission to view or download the files.

    Important Notes

    • Only one Jupyter Notebook should be submitted per group.
    • Ensure that all notebook cells have been executed so that all outputs are visible. Markers will not re-run your code during marking.
    • Please use the Project Summary Report Template, available on:
      WebCMS3 → Project → Project Report Template

    Teamwork

    This is a group project. The aim is for all team members to work together towards a common outcome. Your presentation and submission should reflect a collaborative effort rather than individual work.

    Thanks,
    COMP9444 Admin Team

  • COMP9444 Group Project Presentation Guidelines

    Posted by Fatemeh Barazesh Morgani Monday 09 November 2026, 09:43:21 AM.

    Dear students,

    As we approach Week 10, please note the following important points regarding your group project presentations:

    • Your final presentation will take place during the tutorial session assigned to your team. Please ensure your team attends your scheduled tutorial session to present your project.
    • Each team will have a maximum of 15 minutes, including:
      • 12 minutes for the presentation
      • 3 minutes for Q&A
    • It is essential that all group members attend the final presentation. Since the project includes both individual and group-based marks, students who do not attend will lose marks.
    • Two tutors will assess your presentation using the marking rubric, and your final presentation mark will be the average of the two scores.
    • Live presentations are required. All group members must be present during both the presentation and Q&A.
    • Please use the project presentation template, which is available in WebCMS:
      WebCMS3 > Project > Project Presentation Template
    • We highly recommend the following:
      • On the title slide, include the zID and name of each group member in the order of their presentation.
      • On each slide, clearly state the name and zID of the presenter in a footnote.

    Kind regards,
    COMP9444 Admin Team

Upcoming Due Dates

There is nothing due!

Back to top

COMP9444 26T3 (Neural Networks and Deep Learning) is powered by WebCMS3
CRICOS Provider No. 00098G