Research Intern RL & Post-Training Systems

Requirements

• Pursuing a PhD or MS in Computer Science, EE, or a related field (exceptional undergraduates considered)
• Research experience in one or more of:
  • RL or post-training for large models (e.g., RLHF, RLAIF, GRPO, preference optimization)
  • ML systems (inference engines, runtimes, distributed systems)
  • Large-scale empirical ML research or evaluation
• Comfortable with empirical research: designing controlled experiments, interpreting noisy results, drawing principled conclusions
• Strong Python skills for experimentation
• Willingness to modify inference or training systems (C++, CUDA experience is a plus)

Example Research Directions

• Inference-Aware RL & Post-Training: Designing RL or preference-optimization objectives that account for inference cost and structure; studying how inference-time approximations affect learning dynamics; analyzing bias, variance, and stability trade-offs.
• RL-Centric Inference Systems: Developing inference mechanisms for deterministic, reproducible RL rollouts at scale; exploring batching, scheduling, and memory-management strategies optimized for RL workloads; investigating KV-cache policies and runtime abstractions.
• Scaling Laws & Cost–Quality Trade-offs: Empirically characterizing how reward improvement and generalization scale with rollout cost, latency, and throughput; quantifying when systems-level optimizations change algorithmic behavior.
• Evaluation & Measurement: Designing rigorous benchmarks and diagnostics for post-training and RL efficiency; studying failure modes in long-horizon training.

Preferred Qualifications

• Publications at leading ML and NLP conferences (NeurIPS, ICML, ICLR, ACL, EMNLP)
• Understanding of model optimization techniques and hardware acceleration approaches
• Contributions to open-source machine learning projects

Internship Program Details

• Fall internship program spans 12 to 16 weeks (September 14th to December 18th)
• Opportunity to work with industry-leading engineers and contribute to influential open source projects

Compensation & Benefits

• Competitive compensation, housing stipends, and other competitive benefits
• Estimated US hourly rate: $58 to $63 per hour