Research Intern, Agent RL Training

📋 Description

• Collaborate with mentor to identify high-impact LLM research directions.
• Independently run end-to-end SFT experiments on LLM-based agents.
• Assist with RL exploration: reward design and training iterations.
• Curate high-quality training datasets (instruction-following, agent trajectories, synthetic data).
• Contribute to public publications; support top-venue submissions during internship.

🎯 Requirements

• Highly motivated and able to put in extra hours as needed.
• Genuine passion for research; read papers and tinker with models.
• Independently capable of end-to-end model SFT; basic RL post-training methods (RLHF, DPO, PPO, GRPO).
• Excellent taste in model behavior; able to reason about what good looks like.
• Strong Python and PyTorch skills.

🎁 Benefits

• Mentor pairing with a full-time engineer.
• Opportunity to contribute to top-tier publications during internship.
• Hands-on experience with LLMs, agent RL, and NewsBreak products.
• Collaborative, fast-paced research-focused team culture.