I am an AI Research Scientist at Meta Superintelligence Lab, based in Menlo Park, California. My work centers on media understanding and generation — building foundation models that perceive, reason about, and create images and video at scale. Recent work includes reference-based image generation (Imagine Yourself, AI Sticker) and video generation (Video Ads Creatives), and movie-grade talking-character synthesis (Mocha, NeurIPS Spotlight). Before Meta, I earned my Ph.D. and M.S. from the University of Southern California, where I studied computer vision, deep learning, and the statistical genomics of gene-expression regulation.

[We are hiring] I’m looking for strong, motivated students for AI Research Intern roles in media generation, world models, and long-context agentic memory. If this sounds like you, reach out at bsunmiracle@meta.com.