LLaVA-series researcher reflects on 8-year journey from Didi intern to VLM builder

liuziwei7 · x · 2026-09-20

Researcher Brian Bo Li shares that his work has joined the CS-xxx course series. He started learning AI via the same series during a Didi internship eight years ago, began his PhD in 2022, trained vision-language models with Chunyuan Li in 2023, and joined the LLaVA-series development in 2024. He says he'll focus on shipping better models going forward.

Original post →

More from Companies & People

Companies & People channel →