Choose timezone
Your profile timezone:
Practical ChatGPT for Academics: Deep Research, Code & LaTeX — Live Demo
→
Europe/Ljubljana
Virtual location (ZOOM)
Virtual location (ZOOM)
,
Description
Description: A 2.5 hour live-demo seminar on practical, high-impact use of ChatGPT in academia: effective prompting and voice tactics, verified Deep Research, programmatic problem-solving (live code — digital filter design), figure creation from text/screenshots, and LaTeX outputs (Beamer slides & posters in Overleaf). Hands-on participation optional.
Detailed description: This seminar provides a comprehensive introduction to using ChatGPT as a versatile academic assistant for the entire research workflow. Participants will see, step by step, how to use ChatGPT to perform structured literature reviews, locate relevant open-source repositories, and organize research findings efficiently. The emphasis is on verifiable, citation-based outputs and critical interaction with the model rather than passive use. Real-world examples will show how Deep Research mode can complement traditional search engines and databases when used responsibly.
A major part of the seminar focuses on practical code generation and execution. Attendees will observe how ChatGPT can design, implement, and run small-scale analytical projects directly inside the chat interface, including numerical simulations and data visualizations. The demonstration will include a digital filter design example, highlighting verification strategies, interpretation of results, and common pitfalls. This section will also touch on multimodal features such as generating figures from text or screenshots, showing how image understanding and textual reasoning combine to support reproducible research workflows.
Finally, the seminar will demonstrate how ChatGPT can support academic writing and dissemination. Participants will learn how to automatically produce Beamer presentations and scientific posters directly from LaTeX papers in Overleaf, ensuring consistency between research outputs and presentations. The session concludes with a short bonus demo of text-driven 3D scene generation to illustrate creative applications of language-to-visual synthesis. Throughout the event, best practices for prompting, critical review, and AI literacy will be emphasized to ensure reliable and ethical use of generative models in research and education.
Date and time: 22. October 2025 from 16.00 - 18.30
Language: English
Level: Intermediate
Location: Virtual (Zoom) - link will only be available to registered participants
Target audience: Researchers, graduate and undergraduate students, lecturers, and technical staff in academia who want to integrate ChatGPT effectively into research, coding, writing, and presentation workflows.
Max. number of participants: 300
Prerequisites (optional):
• Overleaf account;
• ChatGPT (Plus/Team/Pro)
Workflow: Live Zoom session with screen-shared demonstrations. The seminar begins with prompting and voice tactics, followed by Deep Research for literature and repository discovery, live Python code execution (digital filter example), figure generation from text or screenshots, and creation of Beamer presentations and posters in Overleaf. Concludes with a short bonus demo (text-described 3D scene) and Q&A.
Learning outcomes:
• Effective prompting and voice/dictation for complex academic tasks.
• Verified literature and code-repository search.
• Using ChatGPT’s built-in code execution for analysis and visualization.
• Generating publication-quality figures from text or screenshots.
• Creating Beamer presentations and posters from LaTeX papers in Overleaf.
• Minimizing hallucinations and improving response reliability.
Organizator:
Lecturers:
| Assoc. prof. dr. Janez Perš |
|
Associate professor at the Faculty of Electrical Engineering, University of Ljubljana, and a member of the Laboratory for Machine Intelligence (LMI). His research and teaching span computer vision, machine intelligence, and embedded systems, with a strong focus on applying artificial intelligence to real-world problems such as autonomous vehicles, robotics, and precision agriculture. He has been an early and active adopter of AI-assisted tools in research, teaching, and academic communication, exploring how systems like ChatGPT can augment productivity, creativity, and critical thinking in academic contexts. |
| janez.pers@fe.uni-lj.si |
| as. dr. Janez Križaj |
| Dr. Križaj is a postdoctoral researcher at the Laboratory for Machine Intelligence, Faculty of Electrical Engineering, University of Ljubljana. His research focuses on computer vision, biometrics, and deep learning. |
| janez.krizaj@fe.uni-lj.si |
Contact
Registration
Registration
