The Department of Computer Science and Engineering in association with Computer Society of India and Software Development & Information Management club organized a Guest Lecture on Technology That Powers CHATGPT (Introduction to LLMS) on 13.05.2024. The objective of this Session was to provide students with a comprehensive understanding of the underlying principles, architecture, and functionalities of large language models (LLMs) such as ChatGPT.
The welcome address was given by S.Divyasri, III Year/CSE. The presidential address was given by Dr.M.Premkumar, Principal, Paavai Engineering College. He emphasized the importance of ChatGPT as a virtual companion ever prepared to participate in meaningful discussions, offer valuable insights, or simply provide friendly conversation, all while enhancing employability in a swiftly changing technological environment.
Mr.K.R.Srikanth Doss, Senior Applied Scientist, Seattle, Washington, USA, commenced the session with a historical perspective on language models, tracing their evolution from simple statistical models to complex neural networks. He explained the basics of language models, focusing on their ability to predict and generate human-like text. He demonstrated how these components work together to enable LLMs to understand and generate coherent text.
The process of training LLMs involves using vast datasets to teach the model to recognize patterns in language. Different training techniques, including supervised learning, unsupervised learning, and fine-tuning, were detailed. Real-world examples were provided to illustrate the versatility and practical uses of LLMs across different industries. The lecture concluded with an interactive Q&A session, where attendees asked questions about various aspects of LLMs. The speaker provided detailed answers, offering deeper insights into the topics discussed. Key points of interest included the limitations of current LLMs, potential improvements, and interdisciplinary applications of AI research.
As a conclusion, the guest lecture was highly informative and well-received, offering a comprehensive understanding of the technology that powers ChatGPT and other LLMs.
The vote of thanks was delivered by J.Bharathkrishnan, III Year/CSE. A total number of 462 students and the faculty members attended this Briefing Session.
