Selected Projects
Teach M-Powered: Using Generative AI to Assist Teachers with Writing Effective Feedback
Stanford University
This project is to investigate the capabilities of Large Language Models (LLMs) to enhance their abilities in crafting feedback for students, aiding teachers in writing effective feedback. Through extensive literature review, I identified various dimensions of effective feedback and formulated a rubric to steer the feedback generation process within LLMs. Additionally, I spearheaded the frontend development of a custom-designed learning management system, integrating insights gained from user research to ensure the tool meets the needs of teachers and addresses existing gaps in the feedback writing process.
M-Powering Teacher: A Machine Learning Tool for Instructional Measurement and Feedback
Stanford University
Through this project, our team has developed a machine learning tool to analyze diverse facets of a teacher's instruction, delivering a holistic report on classroom interaction. My responsibility was to design and develop intuitive data dashboards that effectively communicate the insights gleaned from these interactions. Collaborating closely with teachers and educational researchers, I identified specific feedback needs aimed at elevating classroom practices. This entails designing interfaces that distill complex data into actionable insights, facilitating informed decision-making for teachers and teaching coaches.
Code In Place: A free, human-centered, intro-to-coding course from Stanford University
Stanford University
I worked closely with Professors Dora Demszky and Chris Piech to build the AI-Feedback feature for Code In Place, a large-scale human-centered coding course offered by Stanford, to provide automated feedback to section leaders and students leading to improvement in learning outcomes.
CRAFT: Empowering students with AI Literacy
Stanford University
Through CRAFT, I researched and developed a learning tool that could help students utilize Large Language Models in a critical way and effective way in their writing processes. The tool was developed through a co-design process with students and teachers. While the co-design process offered insights into the unique needs of the students and teachers, the learning sciences literature helped in shaping the tool for pedagogical effectiveness.
Publications
From Consumers to Critical Users: Prompty, an AI Literacy Tool For High School Students
Deepak Varuvel Dennison,* Raycelle C. C. Garcia,* Parth Sarin, Jacob Wolf, Christine Bywater, Benjamin Xie and Victor R.Lee
AAAI-24: Proceedings of the AAAI Conference on Artificial Intelligence
Co-designing AI Education Curriculum with Cross-Disciplinary High School Teachers
Benjamin Xie, Parth Sarin,* Jacob Wolf,* Raycelle C. C. Garcia, Victoria Delaney, Isabel Sieh, Anika Fuloria, Deepak Varuvel Dennison, Christine Bywater, Victor R. Lee
AAAI-24: Proceedings of the AAAI Conference on Artificial Intelligence
Evaluating LLMs’ Generation of Growth-Mindset Supportive Teacher Language
Margarett Clapper, Cameron Hecht, Michaela Jones, Nirel Jonesmitchell, Meghann Johnson, Deepak Dennison, Dorrotya Demszky and David Yeager
AIED2023 Workshop: Equity, Diversity, and Inclusion in Educational Technology Research and Development