Exploring the Integration of Large Language Models with Reinforcement Learning


Are you interested in combining Large Language Models (LLMs) with Reinforcement Learning (RL)? This project/thesis aims to explore innovative ways to enhance RL systems by integrating advanced language models, potentially through reward shaping and human feedback. Key Areas of Focus: LLMs in RL: Investigate how LLMs can be used to improve decision-making and learning in RL. Reward Shaping: Examine techniques for modifying reward signals to boost learning efficiency and performance. Human Feedback: Explore methods for incorporating human guidance into RL to refine and accelerate learning processes. This project is suitable for students eager to contribute to AI research by blending theoretical concepts with practical applications. Please note that the project will be assigned only if I have the capacity to supervise, and the exact topic will be finalized after discussion to align with both the student's interests and research goals.


Python programming, previous experience with ML. Ideal: Experience with RL and LLMs.

Contact person: Mohit Jiwatode