Research Intern (Model Shaping, Winter 2027)
Together AI
- Amsterdam, Netherlands
- Internship
- Intern/New Grad
- Bachelor's degree
- Posted
Skills
Required
- Machine Learning
- Deep Learning
- Transformer architecture
Preferred
- deep learning frameworks
- PyTorch
- JAX
- training foundation models
- ML
- NLP
- NeurIPS / ICML / ICLR / ACL / EMNLP
- open-source
- machine learning projects
- post-training
Join Together AI as a Research Intern in the Model Shaping team. You will work on advanced post-training methods, efficient training techniques for neural networks, and robust evaluation of foundation model capabilities. This internship offers the opportunity to conduct rigorous experiments, document findings in scientific publications, and integrate research results into Together products. The internship lasts 12 to 14 weeks, with dates spanning January 4th to April 9th, and is available in either the San Francisco or Amsterdam office.
Understanding of model optimization techniques and hardware acceleration approaches
Responsibilities
- Research and implement novel techniques in advanced post-training methods, efficient training of neural networks, and robust evaluation of foundation model capabilities.
- Design and conduct rigorous experiments to validate hypotheses, document findings in scientific publications and blog posts, and integrate research results into Together products.
- Collaborate with industry-leading engineers and researchers across multiple teams to enhance machine learning models and contribute to the development of new methods for model training and evaluation.
Requirements
- Familiarity with the Transformer architecture and recent developments in foundation models
- Strong knowledge of Machine Learning and Deep Learning fundamentals
- Currently pursuing a Bachelor's, Master's, or Ph.D. degree in Computer Science, Electrical Engineering, or a related field
Preferred
- Experience with deep learning frameworks (PyTorch, JAX, etc.)
- Prior research experience with training foundation models or efficient machine learning
- Publications at leading ML and NLP conferences (such as NeurIPS, ICML, ICLR, ACL, or EMNLP)
- Contributions to open-source machine learning projects
