Meta
Research Scientist Intern, FAIR - Video Generation Self-Improvement (PhD)
Menlo Park, CA +1 · Internship · Posted Oct 27, 2026
- Location
- Menlo Park, CA
Seattle, WA - Education
- Currently has or is in the process of obtaining a PhD degree in Computer Science, Electrical Engineering, Computer Vision, Machine Learning, Artificial Intelligence, relevant technical field, or equivalent.
- Visa sponsorship
- Must obtain work authorization in the country of employment by the time of hire and maintain it during employment.
About the job
Conduct research and develop algorithms, prototypes, and experiments for video generation, multimodal model post-training, and self-improving agentic systems.
What you'll do
- Research video generation, video synthesis, and multimodal learning; develop and optimize post-training methods for large-scale models.
- Develop self-improving and agentic generation systems, including multi-agent LLM/VLM pipelines and reward modeling.
- Implement and evaluate research prototypes, analyze results, and communicate findings through reports, publications, and presentations.
What they ask for
- Currently has or is pursuing a PhD in a relevant technical field, or equivalent.
- Research or work experience in video generation, computer vision, or multimodal AI.
- Experience with large-scale model training, post-training optimization, and/or data curation.
- Experience with Python, C++, or related languages and deep learning frameworks such as PyTorch or TensorFlow.
Nice to have
- Experience with video-language models, LLM/VLM agents, reward modeling, or generative-model evaluation.
- Research achievements such as publications at leading conferences, journals, or workshops, or grants, fellowships, or patents.
Pay
$7,650/month to $12,667/month (base salary)
Tools
Python, C++, PyTorch, TensorFlow, Llama, LLM, VLM
AI focus
Video generation and synthesis; multimodal model post-training using reinforcement learning from human and AI feedback; reward modeling; and self-improving, agentic systems.
Model-written summary · the original posting is authoritative