Probabilistic Foundation Model Research Intern

DataRobot; Washington (District of Columbia); On-site; USD 60,000 - 90,000; Full time

Benefits offered by this job

Medical Insurance

Dental Insurance

Vision Insurance

Paid Time Off

Parental Leave

EAP program

Job summary

DataRobot seeks a research‑engineering intern at post‑doctoral level to shape architecture of probabilistic foundation models. You will implement, train, and evaluate PyTorch models in a production‑quality codebase, focusing on temporal, tabular, and mixed‑modality data with attention mechanisms and distributional heads.

The role emphasizes precise reasoning about architectural capabilities, controlled architecture studies, and scalable PyTorch implementations within a fast, rigorous research

Qualifications

• Post‑doc level research internship focusing on probabilistic foundation models.
• Strong PyTorch skills and hands‑on experience building and training deep models.
• Experience with Transformers and long‑context architectures; memory and latency tradeoffs.
• Background in stochastic modeling is a strong plus, not a prerequisite.
• Ability to implement, train, and evaluate production‑quality PyTorch code.

Responsibilities

• Design and improve core model architecture for temporal and mixed‑modality data.
• Run controlled architecture studies, ablations, and scaling experiments.
• Build scalable PyTorch implementations with a focus on memory budgets and throughput.
• Study how data generation and inference constraints affect benchmark quality and real‑world usefulness.
• Turn research ideas into robust implementations and credible results.

Skills

PyTorch

Transformers

Probabilistic modeling

Probability & statistics

GPU programming

Experimentation

Debugging & testing

Education

PhD or near PhD in quantitative field

Job description

DataRobot seeks a research‑engineering intern at post‑doctoral level to shape architecture of probabilistic foundation models. You will implement, train, and evaluate PyTorch models in a production‑quality codebase, focusing on temporal, tabular, and mixed‑modality data with attention mechanisms and distributional heads.

The role emphasizes precise reasoning about architectural capabilities, controlled architecture studies, and scalable PyTorch implementations within a fast, rigorous research