Ask HN: Does Attention (from LLMs) work well on any other data types?

Just wondering if theres any other research out there of applying the attention mechanism from large language models to any other data types (spitballing but like weather data, or IoT data, etc). If there is research on it, would you mind providing a few citations?