Accelerated Science Intern, Representation Learning for Molecules and Materials

What you'll do

• Implement and evaluate models using reproducible workflows; design experiments comparing representations across training runs, architectures, and learning objectives.
• Analyze representation geometry and information content, and develop methods to identify meaningful shared structure.
• Contribute documented code and communicate findings through reports, publications, or patent disclosures.

What they ask for

• Currently enrolled in a doctoral program in a relevant discipline.
• Strong foundations in machine learning, linear algebra, probability, and optimization, with demonstrated experience training models on real datasets.
• Proficiency in Python and modern ML tooling.
• Ability to collaborate across disciplines and translate research ideas into working code.

Nice to have

• Familiarity with representation learning or scientific machine learning.
• Experience studying model convergence, learning dynamics, symmetries, or identifiability.

Pay

$45 and $65/hour (USD), base pay

Tools

Python, PyTorch

AI focus

Machine-learning research on learned representations of molecules and materials, including their shared structure, information content, and relationship to physical quantities.