Instructor: Sungwoong Kim
TA: TBD
Time: Fri 13:30 - 16:15 (break time: 14:45 - 15:00)
Room: Woojung B102
Contact: swkim01@korea.ac.kr or blackboard
Large language models (LLMs) has now expanded beyond their roles in natural language processing and are being used as foundational models for implementing AGI, not only in language but across all domains and modalities. Of course, through natural language processing, AI agent based on LLMs can communicate deeply and tightly with a human, leading to a massive amount of real applications. This course will focus on various cutting-edge researches based on LLMs including Multimodal LLMs, Reasoning LLMs, recent Human Alignment algorithms.
Throughout the course, students will comprehensively understand key factors in cutting-edge algorithms for recent natural language processing especially based on LLMs. In particular, this course will discuss and delve deeply into the latest papers and research studies on LLM-based natural language processing including Reasoning LLM, ToolLLM, advanced human alignment algorithms, Efficient LLM, Multi-Modal LLM, Non-Autoregressive LLM, VLA, and Mixture-of-Experts LLM. Eventually, the course aims students to develop practical knowledge and skills and moreover to have an insight for latest researches to leverage those for their own research.
Basic knowledges in machine learning, deep learning, natural language processing, generative modeling, and reinforcement learning are required.
Lecture notes will be the main material of the course, and these do not come from a single textbook.
Participation (20%)
Attendance (20%)
Midterm Exam (30%)
Final Exam (30%)