About the role#
We are looking for a Research Intern to join our Agent RL Training team. You will be paired with a full-time mentor to explore how to apply large language models to core business areas, including content understanding, recommendation, agentic web browsing, and autonomous multi-step task completion. This is a hands-on role where you will independently drive experiments, propose novel ideas, and iterate quickly.
What you'll do#
- Collaborate with your mentor to identify high-impact research directions for applying LLMs to products.
- Independently run end-to-end SFT experiments on LLM-based agents and assist with RL-related exploration, such as reward design and training iteration.
- Curate and build high-quality training datasets, including instruction-following, preference pairs, agent trajectories, and synthetic data.
- Contribute to public publications, with support for top-venue submissions during your internship.
What you'll need#
- Strong Python and PyTorch skills.
- Independent capability in end-to-end model SFT, with a basic understanding of RL-based post-training methods such as RLHF, DPO, PPO, or GRPO.
- Excellent taste in model behavior, with the ability to reason about quality across user-facing domains.
- Genuine passion for research, demonstrated by reading papers and tinkering with models.
- High motivation and commitment, including the willingness to put in extra hours to push projects to completion.
- Pursuing a degree in Computer Science, Data Science, or Mathematics.
- Preferred: Publication at a top-tier venue, experience with multi-node distributed training, proficiency in writing custom GPU kernels with Triton or CUDA, experience building synthetic data pipelines, and familiarity with open-source RL frameworks.
Location & details#
- Location: On-site in Mountain View, California.
- Type: Full-time internship.
- Compensation: $35 - $50 per hour.
- Sponsorship: Not provided.
About NewsBreak
NewsBreak operates a digital platform that aggregates local news and community information. The company connects users with content from over 10,000 sources across the United States. It maintains a focus on machine learning and natural language processing to personalize information delivery. Founded in 2015, the organization is based in Mountain View, California.
How to get in at NewsBreak
Applying early to NewsBreak is a significant advantage because recruiters often review the first wave of candidates before the applicant pool becomes unmanageable. Intern Insider sends an instant alert the moment a role matching your target is published, which helps you submit your materials among the first. This timing ensures your profile is visible while the team is still actively building their shortlist. You can also improve your odds by reaching out to the people actually hiring for the team. Intern Insider surfaces the recruiters behind NewsBreak roles so you can contact them directly to ask about the position or request a referral. A brief, professional message to the right person often yields a better response rate than waiting for an automated system to process your application.
