Researcher, Multimodal Safety

Researcher on the Chat and Multimodal Safety team at OpenAI. This role focuses on advancing multimodal safety research for text, vision, and audio, and building training and evaluation methods for VLMs. The researcher will collaborate with other teams to translate research into safer multimodal experiences.

Skills and focus areas

• multimodal safety research
• vision-language models
• modality fusion
• image encoders
• multimodal post-training
• multimodal evaluations
• safety for image systems
• safety for video systems
• safety for audio systems
• perception
• semantic understanding
• training methods