Accelerated Science Intern, Representation Learning for Molecules and Materials

Toyota Research Institute•Los Altos, CA

4d•$45 - $65•Hybrid

About The Position

Toyota Research Institute (TRI) is hiring a paid research intern for a 12-week, hybrid, in-office internship. TRI develops tools and capabilities to improve human life and advance mobility through work in AI, robotics, driving, and materials science. The intern will join TRI’s Accelerated Materials Design and Discovery (AMDD) program, whose long-term vision is to accelerate the development of emissions-free mobility by discovering materials and devices for batteries, fuel cells, and other applications. The program combines computational materials modeling, experimental data, artificial intelligence, and automation, and collaborates with universities, national laboratories, and Toyota colleagues worldwide. The research focuses on developing algorithms and computational systems for multi-agent learning and coordination. The intern will investigate how agents with different data and computational capabilities can share information, build predictive models, and select computational tasks to improve collective understanding. Potential approaches include distributed and federated learning and information-theoretic methods for communication and task selection. The intern will implement and benchmark methods, evaluating effects on model accuracy, communication costs, and computational efficiency.

Requirements

Currently enrolled in a doctoral program in computer science, machine learning, statistics, applied mathematics, electrical engineering, or a related discipline.
Strong foundations in machine learning, probability, and optimization, with demonstrated experience implementing and evaluating algorithms.
Familiarity with multi-agent systems, distributed or federated learning, or information theory.
Proficiency in Python and modern machine-learning tooling, such as PyTorch, experiment tracking, and reproducibility best practices.
Ability to work collaboratively across disciplines and translate research ideas into working code.

Nice To Haves

Experience with distributed computing.

Responsibilities

Implement and evaluate algorithms for learning and coordination among agents with different data and computational capabilities.
Prototype methods for sharing information, updating predictive models, and selecting computational tasks.
Design controlled benchmarks comparing coordination strategies with independent-agent and centralized approaches.
Evaluate model accuracy, communication costs, and computational efficiency using reproducible experimental workflows.
Contribute high-quality, well-documented code to an internal codebase.
Help translate results into internal reports, publications, and/or patent disclosures, as appropriate.