Exploring
VLM
Vision-language models. Exploring: vocabulary and reading, no public fine-tune.
Simple explanation
A VLM ties images (and sometimes video) to language. That is not yet a robot.
Technical explanation
A public pixel baseline exists (UCI digits 8×8, majority vs centroid). That is not a VLM. PhotoMind stays private.
Evidence
Public pixel baseline: the 8×8 digits lab. No fine-tuned VLM. See also the Physical AI lab (vocabulary map).