Research Intern, Agent RL Training

Responsibilities

• Collaborate with a mentor to identify high-impact research directions for applying LLMs to NewsBreak products
• Independently run end-to-end SFT experiments on LLM-based agents and assist with RL exploration including reward design and training iteration
• Curate and build high-quality training datasets: instruction-following, preference pairs, agent trajectories, and synthetic data
• Contribute to public publications and top-venue submissions

Requirements

• Highly motivated and committed
• Passion for research and deep intellectual curiosity
• End-to-end model SFT capability
• Basic understanding of RL-based post-training methods (RLHF, DPO, PPO, GRPO)
• Excellent taste in model behavior
• Strong Python and PyTorch skills

Preferred Qualifications

• Publication at a top-tier venue (NeurIPS, ICML, ICLR, ACL, EMNLP, or equivalent)
• Experience with multi-node distributed training (FSDP, DeepSpeed, Megatron-LM)
• Proficiency in writing custom GPU kernels with Triton or CUDA
• Experience building synthetic data pipelines for agent training
• Familiarity with open-source RL frameworks: TRL, OpenRLHF, veRL/vLLM

About the Company

NewsBreak is a Content Intelligence platform founded in 2015 with over 40 million monthly active users. It delivers personalized local news and information powered by advanced AI, recommendation systems, and adtech. The company is Great Place to Work certified and reached unicorn status in 2021.