- Company Name
- Wayve
- Job Title
- Research Scientist Intern, Embodied Foundation Models (Evaluation)
- Job Description
-
**Job Title**
Research Scientist Intern – Embodied Foundation Models (Evaluation)
**Role Summary**
A 3‑6 month internship for a graduate student focused on training, evaluating, and advancing embodied AI foundation models. The role involves large‑scale multimodal (vision‑language) pre‑training, distributed training on multi‑node systems, and benchmarking. The intern will spearhead a research project, analyze results, and aim to publish findings at top AI/robotics conferences.
**Expectations**
- Deliver end‑to‑end training and evaluation pipelines for multimodal foundation models.
- Produce reproducible results and diagnostics for open and proprietary datasets.
- Write high‑quality research manuscripts for conferences such as CVPR, ICCV, NeurIPS, CoRL, RSS, or ICRA.
- Collaborate with applied scientists, ML engineers, and software engineers to integrate research insights into products.
**Key Responsibilities**
1. Design and implement distributed training procedures for large vision‑language models.
2. Optimize data pipelines for multimodal datasets and evaluate performance on benchmark suites.
3. Conduct ablation studies, compare reasoning capabilities, and interpret results.
4. Document experiments, maintain reproducibility, and share findings with the team.
5. Author and co‑author conference papers, presenting results at internal and external meetings.
**Required Skills**
- Advanced programming in Python; experience with backend/systems languages (e.g., Ruby, Java) is a plus.
- Proficiency with deep‑learning frameworks: PyTorch, TensorFlow, or JAX.
- Hands‑on experience with large‑scale distributed training (multi‑node, GPU/TPU clusters).
- Strong background in vision‑language models, large language models, or NLP with reasoning focus.
- Familiarity with benchmarking tools, metric analysis, and reproducibility best practices.
- Ability to translate research insights into clear, technical documentation and presentations.
**Required Education & Certifications**
- Currently enrolled in a graduate program (MS/PhD) in Computer Science, Machine Learning, Robotics, or a closely related field.
- Submission record of peer‑reviewed work in venues such as CVPR, ICCV, NeurIPS, CoRL, RSS, or ICRA.
- No specific certifications required, but a proven track record in AI research and software development is essential.