Date
Publisher
arXiv
This paper presents the ADAIO team's system entry in the Building Educational
Applications (BEA) 2023 Shared Task on Generating AI Teacher Responses in
Educational Dialogues. The task aims to assess the performance of
state-of-the-art generative models as AI teachers in producing suitable
responses within a student-teacher dialogue. Our system comprises evaluating
various baseline models using OpenAI GPT-3 and designing diverse prompts to
prompt the OpenAI models for teacher response generation. After the challenge,
our system achieved second place by employing a few-shot prompt-based approach
with the OpenAI text-davinci-003 model. The results highlight the few-shot
learning capabilities of large-language models, particularly OpenAI's GPT-3, in
the role of AI teachers.
Study at a glance
- Question: How well can artificial intelligence generate appropriate teacher responses during conversations with students?
- Setting and data: Student-teacher dialogues from the 2023 Building Educational Applications (BEA) Shared Task.
- Method: Evaluation of AI models using different prompt designs to generate teacher responses (few-shot prompt-based approach)
Why it matters
Using generative models as AI teachers requires assessing their performance in producing suitable responses within a student-teacher dialogue.
Key findings
- In an educational dialogue competition, an AI system achieved second place for generating suitable teacher responses to students.
- Large language models, especially GPT-3, demonstrated strong capabilities to serve as AI teachers within student conversations.
- Generative AI models were able to learn to produce teacher responses effectively when shown only a few examples.
Summary drafted with AI from the paper’s abstract and may contain errors. See the full paper for details.
What is the application?
Who is the user?
Why use AI?
Study design
