Research Scientist Graduate - (Trust and Safety - CV/NLP/Multimodal LLM) - 2026 Start (BS/MS)

Join the Trust and Safety algorithm team to develop computer vision and multimodal models that detect violating content and behaviors. You’ll explore cutting-edge multimodal foundation models and apply LLM techniques such as pre-training, zero-shot/few-shot learning, and hard case mining. You will also continuously improve the training framework so large models perform more effectively for trust and safety scenarios.