Research Intern: Foundation Models & ML Systems (12-14 wks)

Together Computer Inc; Amsterdam; On-site; EUR 17,000 - 23,000; Full time

Benefits offered by this job

Health insurance

Dental insurance

Vision insurance

401(k) plan

Commuting stipend

Job summary

Together AI's Model Shaping team in Amsterdam invites a Research Intern to work on advanced post-training methods, distributed training techniques, and robust evaluation of foundation models. You will design and run experiments across supervised learning, reinforcement learning, and preference optimization, and contribute to scientific publications.

The 12 to 14 week internship spans January to April, offering collaboration with engineers and researchers across multiple teams and opportunities

Qualifications

• Familiarity with Transformer architecture and foundation models developments.
• Strong knowledge of ML and DL fundamentals.
• Currently pursuing a Bachelor’s, Master’s, or Ph.D. in CS, EE, or related field.
• Experience with deep learning frameworks (PyTorch, JAX, etc.).
• Prior research in training foundation models or efficient ML.
• Publications in leading ML/NLP conferences (NeurIPS, ICML, ICLR, ACL, EMNLP).
• Contributions to open-source ML projects.
• Understanding of model optimization techniques and hardware acceleration approaches.

Responsibilities

• Design and conduct rigorous experiments to validate hypotheses.
• Research and implement novel techniques in focus areas.
• Document findings in scientific publications and blogs.
• Integrate results into Together products.

Skills

Transformer familiarity

ML & DL fundamentals

Pursuing degree (BSc/MSc/PhD) in CS/EE

Deep learning frameworks experience

Foundation models research

Publications in ML/NLP venues

Open-source contributions

Model optimization concepts

Education

Bachelors/Masters/PhD in CS or EE

Tools

PyTorch

JAX

Job description

Together AI's Model Shaping team in Amsterdam invites a Research Intern to work on advanced post-training methods, distributed training techniques, and robust evaluation of foundation models. You will design and run experiments across supervised learning, reinforcement learning, and preference optimization, and contribute to scientific publications.

The 12 to 14 week internship spans January to April, offering collaboration with engineers and researchers across multiple teams and opportunities