Research Intern – Applied Reinforcement Learning

This is a fully remote position, open to applicants in California, +1 more state.

📋 Description

• Design and assess reinforcement learning (RL) systems tailored for agentic AI workflows.

• Develop RL environments, reward models, and post-training pipelines specifically for LLM-based agents.

• Construct comprehensive RL pipelines for agentic systems encompassing simulation, training, and evaluation.

• Align LLM-based agents utilizing RLHF, DPO, PPO, and other innovative methods.

• Create reward functions, verification processes, and evaluation frameworks.

• Develop simulation environments (digital twins) to support enterprise workflows.

• Guarantee scalable training and inference capabilities for RL-based systems.

• Document experiments, ablations, and results for both research and production purposes.

⛳️ Requirements

• PhD candidate in Computer Science, Machine Learning, or a related discipline with a focus on reinforcement learning or agentic AI.

• Proficiency in Python and PyTorch, along with experience in GPU-based training.

• Strong grasp of RL fundamentals, including MDPs, policy gradients, and value methods.

• Familiarity with LLMs and post-training methodologies such as RLHF, DPO, and PPO.

• Excellent experimentation skills, including ablation studies, reproducibility, and clear documentation.

• Experience with RL environments such as Gymnasium, RLlib, or Stable Baselines is preferred.

• Research background in offline RL, model-based RL, or hierarchical RL is preferred.

• Publications in leading ML conferences such as NeurIPS, ICML, ICLR, or ACL are preferred.

• Knowledge of simulation, synthetic data, or multi-agent systems is preferred.

• Experience with distributed training and large-scale experimentation is preferred.

🏝️ Benefits

• Competitive stipend.

• Mentorship from experienced researchers and engineers.

• Access to cutting-edge GPU infrastructure.

• Opportunities to publish and present research findings.