PhD Research Intern: Reinforcement Learning for LLMs

NVIDIA Corporation; Santa Clara (CA); On-site; USD 67,000 - 100,000; Full time

Job summary

NVIDIA Corporation in Santa Clara is seeking PhD research interns to push the next generation of large language models through reinforcement learning. You will design, implement, and evaluate RL-based methods, with a hands-on focus on rapid prototyping at scale.

Work involves developing algorithms for LLM reasoning, alignment, and multi-turn interaction, with experiments on large GPU clusters using Python and PyTorch. Publications or open-source contributions are encouraged.

Qualifications

• Pursuing a PhD in AI, ML, CS, CE, EE, Math, Physics, or a related field.
• Strong background in reinforcement learning and natural language processing.
• Excellent programming skills, especially in Python.
• Experience with deep learning frameworks such as PyTorch.
• Comfort with experimental research and large-scale training pipelines.

Responsibilities

• Develop and prototype reinforcement learning algorithms for large language models.
• Explore methods to improve reasoning, alignment, instruction following, and multi-turn interaction.
• Design experiments to evaluate model behavior, robustness, and task performance.
• Implement research ideas in Python and PyTorch, and run experiments on large-scale GPU clusters.

Skills

Python

Reinforcement learning

Natural language processing

PyTorch

Experimentation

Research pipelines

Education

PhD in AI/ML/CS/CE/EE/Math/Physics

Tools

PyTorch

Git

Job description

NVIDIA Corporation in Santa Clara is seeking PhD research interns to push the next generation of large language models through reinforcement learning. You will design, implement, and evaluate RL-based methods, with a hands-on focus on rapid prototyping at scale.

Work involves developing algorithms for LLM reasoning, alignment, and multi-turn interaction, with experiments on large GPU clusters using Python and PyTorch. Publications or open-source contributions are encouraged.