Research Scientist Intern (Trust and Safety - CV/NLP/Multimodal LLM) - 2027 Start (PhD)

What you'll do

• Develop computer vision or multimodality models to recognize violating content.
• Explore multimodal and computer vision large models and apply LLMs to business scenarios, including pre-training, zero-shot/few-shot learning, and hard case mining.
• Optimize training frameworks for large-model training.

What they ask for

• PhD candidate in computer science or a related technical discipline.
• Research experience in at least one of computer vision, multimodality, or LLM.
• Proficiency with at least one deep learning framework, such as PyTorch or TensorFlow.
• Excellent analytical, problem-solving, logical thinking, communication, and collaboration skills.

Nice to have

• Papers published in top AI conferences or journals, such as CVPR, ICCV, ECCV, NIPS, ICML, ICLR, TPAMI, or IJCV.

Tools

CLIP, COCA, ALBEF, BLIP, Flamingo, ViT-G, ViT-22B, EVA-enormous

AI focus

Research and develop computer vision, multimodal, and LLM models for trust and safety, including violation detection, LLM business applications, and large-model training optimization.