Developing open-source domain-specific chatbots with multimodal capabilities to empower students with interactive question sessions and revolutionize exam preparation.
Leveraging reinforcement learning from human feedback (RLHF) and image captioning techniques to improve the performance of large language models (LLMs) in answering multimodal physics questions.