3 ms·Vision language models. Basically an LLM plus a vision encoder, so the LLM can look at stuff.by pwatsonwailes 1y agoVision language models. Basically an LLM plus a vision encoder, so the LLM can look at stuff.