Research Intern RL & Post-Training Systems, Turbo (Fall 2026)

<h3><strong>About the Role</strong></h3> <p>The <strong>Turbo Research</strong> team investigates how to make <strong>post-training and reinforcement learning for large language models efficient, scalable, and reliable</strong>. Our work sits at the intersection of <strong>RL algorithms</strong>, <strong>inference systems</strong>, and <strong>large-scale experimentation</strong>, where the <strong>cost and structure of inference dominate overall training efficiency</strong> and shape what learning algorithms are practical.</p> <p>As a research intern, you will study RL and post-training methods whose performance and scalability are tightly coupled to <strong>inference behavior</strong>, co-designing <strong>algorithms and systems</strong> rather than treating them independently. Projects aim to unlock new regimes of experimentation—larger models, longer rollouts, and more complex evaluations—by rethinking how inference, scheduling, and training interact.</p> <h3><strong>Requirements</strong></h3> <ul> <li>Pursuing a PhD or MS in Computer Science, EE, or a related field (exceptional undergraduates considered)</li> <li>Have research experience in one or more of:</li> <ul> <li>RL or post-training for large models (e.g., RLHF, RLAIF, GRPO, preference optimization)</li> <li>ML systems (inference engines, runtimes, distributed systems)</li> <li>Large-scale empirical ML research or evaluation</li> </ul> <li>Are comfortable with empirical research by designing controlled experiments, while interpreting noisy results and drawing principled conclusions</li> <li>Can work across abstraction layers:</li> <ul> <li>Strong Python skills for experimentation</li> <li>Willingness to modify inference or training systems (experience with C++, CUDA, or similar is a plus)</li> </ul> </ul> <h3><strong>Example Research Directions</strong></h3> <p>Intern projects are tailored to your background and interests, and may include:</p> <ul> <li><strong>Inference-Aware RL &amp; Post-Training</stro