Neuroscience × AI · Penn AI Fellow
I study how primate brains encode social intelligence and how those principles can inform artificial ones. My work combines neuroscience, machine learning, and animal behavior to model social cognition from neural, video, and audio data — and to probe how artificial systems represent the world.
- Register tokens in vision transformers — Zero-ablation overstates token function in ViTs; under proper controls, register content is fungible. CVPR 2026 (HOW workshop)
- Primate neuro-ethology — Neural signatures of natural behavior in macaques · Nature
- PrimateFace — Cross-species primate facial-analysis benchmark
- awesome-computational-primatology — Curated ML × primatology resource
- Waymo — Safety Research, Human Behavior Analytics · Summer 2026
- Google — Data Science Research · 2024 · built an LLM + human-in-the-loop library for generative-music evaluation



