We can't ask it questions, but we can enter a data sequence and let the model extrapolate it. Similar to how foundation language models (without instruction tun

We can't ask it questions, but we can enter a data sequence and let the model extrapolate it. Similar to how foundation language models (without instruction tuning) used to work, but not just for human readable text.