Research Intern

What you'll do

• Develop research questions and run controlled experiments on datasets, rewards, and training methods.
• Create evaluation datasets, benchmarks, scoring methods, and research tools; analyze model failures and experimental results.
• Communicate findings and collaborate on research publications, benchmark releases, and AI systems.

What they ask for

• Currently pursuing a master’s or PhD in a relevant field.
• Demonstrated ability to formulate research questions, design experiments, and draw sound conclusions from empirical results.
• Experience in at least one area: language-model training, fine-tuning, or evaluation; agentic AI systems or RL environments; or benchmark, evaluation, or data-quality development.
• At least one publication or open-source project; strong programming skills, particularly in Python.

Nice to have

• Previous research in language models, reinforcement learning, model evaluation, or post-training.
• Familiarity with RLVR techniques, reward modeling, or agentic AI systems.

Pay

$80 per hour (base salary)

Tools

Python, GRPO, DAPO

AI focus

Research on post-training, RLVR, data generation, and model evaluation, including reward shaping, language-model capabilities, tool use, agentic behavior, and real-world reasoning.