Blog
Things worth writing down
A mix of research notes, technical explorations, and the occasional essay on things I'm thinking through.
What Cloud Segmentation Taught Us About Vision-Language Models
We tested whether prompt engineering can adapt vision-language models to satellite imagery. Across 60 prompt variants on cloud segmentation, every one underperformed the zero-shot baseline. But fine-tuning with just 8 labeled images beat them all, and the choice between LoRA and full fine-tuning turned out to be about task structure, not compute.
Read postNo posts in this category yet.