Job Description

• Analyze the fundamental limitations of current reinforcement learning (RL) methods for training reasoning models in large language models (LLMs) and vision-language models (VLMs).
• Develop and experimentally validate novel algorithms from a probabilistic inference perspective to address these challenges.
• Collaborate with the team to publish findings in top-tier venues.

Qualifications

• Master’s degree in a relevant field.
• Strong foundation in reinforcement learning, LLMs, VLMs, and probabilistic inference.
• Proficiency in Python and deep learning frameworks (e.g., PyTorch, TensorFlow).
• Proven research record demonstrated by publications in top-tier AI/ML venues.

More Information

Location: Kent Ridge Campus

Organization: School of Computing

Department : Department of Computer Science