Machine Learning Research Intern (Summer 2027)

Scale AI•San Francisco, NY

15h•Onsite

About The Position

Scale’s Machine Learning Research team develops foundational AI systems, spanning generative AI, reinforcement learning and post-training, scalable oversight, synthetic data, red teaming, and evaluation science. The team is building large-scale hybrid human-machine systems for machine-learning pipelines used by customers across industry and government. As an intern, you will work on technical machine-learning applications in production and research problems, with mentorship from researchers and opportunities to collaborate with industry and academic teams. Interns own work that is published or shipped. Example project areas include evaluating frontier-model capabilities and preparedness; creating benchmarks; improving reasoning, reliability, oversight, and post-training methods; studying generalization; scaling human-in-the-loop data generation; improving and deploying production models; and developing agentic or post-training solutions for enterprise and public-sector applications. The internship is available for Summer 2027, with May or June start dates, in San Francisco or New York. Candidates must be enrolled in a relevant BS, MS, or PhD program and have a graduation date in Fall 2027 or Spring 2028. Scale’s team credos include earning customer trust, enabling collaboration across teams, prioritizing consistent quality, focusing on the inputs that drive outcomes, advancing research at the frontier, and planning for downstream consequences. The stated base salary range for this full-time position in San Francisco and New York is $120,000–$150,000 USD. Eligible compensation packages may include equity and benefits. Benefits include comprehensive health, dental, and vision coverage, retirement benefits, a learning and development stipend, generous PTO, and potentially a commuter stipend.

Requirements

Currently enrolled in a BS, MS, or PhD program focused on Machine Learning, Deep Learning, or Natural Language Processing.
Graduation date in Fall 2027 or Spring 2028.
Available for a Summer 2027 internship with a May or June start date in San Francisco or New York.
Prior research experience or a track record of publications in LLMs, NLP, multimodal models, agents, safety, evaluation, alignment, or a related field.
Experience with one or more general-purpose programming languages, such as Python or JavaScript.

Nice To Haves

Previous internship in Machine Learning, Deep Learning, Natural Language Processing, Adversarial Robustness, Alignment, Evaluation, or Agents.
Researcher experience through internships, full-time roles, or at a lab.
Publications at venues such as NeurIPS, ICML, ICLR, CVPR, ACL, EMNLP, or AAAI, or contributions to open-source projects.
Hands-on experience with reinforcement learning, agentic systems, or evaluation infrastructure.

Responsibilities

Conduct research on dangerous capabilities of frontier models and preparedness.
Research and develop benchmarks for frontier models.
Develop training methods for difficult reasoning problems involving long chains of thought.
Research reinforcement learning and post-training methods to improve model reliability on high-stakes tasks.
Research scalable oversight protocols that enable humans to produce and quality-control reasoning chains.
Study model generalization and capabilities to inform data-driven improvements.
Research synthetic and hybrid data generation with humans in the loop.
Identify opportunities to improve production models, retrain and tune them, and deploy improvements without regressing on core model characteristics.
Create post-training or agentic solutions for enterprise and public-sector applications.
Own work that is published or shipped.