Inspiration
Physics is an interesting and useful, but very difficult, subject. I have seen many students struggling in grasping physics concepts. Listening to a teacher and reading a textbook are generally not adequate for learning physics. Learning physics is quite nonlinear and requires multiple literature sources. I found that PartyRock can simplify generating an LLM based app, useful for making physics a fun.
What it does
This is a single-page web app, accessible from an Internet browser. The page gives a brief information about the app, an input box for entering a topic of your interest, output boxes for description, formulation, diagram, and questions, and an interactive box for chats. Description gives an overview of the topic specified by you. Formulation explains the topic through equations. Diagram shows a picture relevant to the topic. Questions offers 3 multiple-choice questions that can be useful for an examination. Chat-box allows you to ask questions and receive AI-generated answers in an interactive mode.
How we built it
PartyRock made the app building a 'piece of cake'. Just after my writing the app objective, the PartyRock framework provided me with a nice front-side layout, supported in full by necessary back-side LLM algorithms. Thereafter, I adjusted the widgets to make them cohesive.
Challenges we ran into
The PartyRock framework had included a widget to show the image of a physicist associated with the topic. I noticed that the framework gave an image of Einstein, irrespective of the specified topic. To avoid confusion, I removed this widget while finalizing my app.
Accomplishments that we're proud of
I was amazed by the potential of PartyRock framework and Large Language Models. I am confident that this app will be immensely useful for students (and even teachers and researchers) scratching their head for unraveling mysteries of physics.
What we learned
I learned that PartyRock can revolutionize the education field through LLM's and AI.
What's next for PhysicsExplorer
I would love to understand the behind-the-scene algorithms used for widgets of this app. If possible, I would make this app useful even to researchers by linking it to research journals.
Built With
- html
- partyrock
- wbem

Log in or sign up for Devpost to join the conversation.