I'm building a dataset of small experiments that test common LLM prompting techniques.

The shared playground links are from Prompt Refine, which means you can fork and run them on your own and try to guess how the LLM will respond.

I'm always adding more experiments, so feel free to submit yours!