Research Scientist, Multimodal Alignment, Safety, and Fairness
Core
Lead breakthrough research in multimodal AI safety, alignment, and fairness, developing novel methods to model AI-society interactions and evaluate evolving AI behaviors.
Role type
Senior Research Scientist (AI Safety & Alignment)
Builds
Novel tools, algorithms, and paradigms for augmented oversight of multimodal AI systems (e.g., Gemini, Nano Banana).
Domain
Artificial General Intelligence (AGI), Multimodal AI, Sociotechnical Modeling
Deliverable
production ML models
Required skills
Deep learning, Computer vision, Generative architectures, Python programming, Experimental design, Independent research
Preferred skills
Large-scale vision language models, VLM prompting strategies, RL-based post-training, Agentic AI solutions, Generative AI techniques, Reinforcement Learning, Multimodal learning
Technologies
JAX, Flax, Gemax
Responsibilities
Generate new ideas and discuss research directions; Design, implement, and evaluate cutting-edge methods and tools; Report and present research findings; Engage in cross-functional collaborations; Take ownership of substantial technical projects from ideation to evaluation.
Seniority
Senior, hands-on IC
