## Prompt Engineering Techs  ##### (GPT3.5 and GPT4) --- ##### (ELEMENTARY LEVEL) # Zero-shot ### Use the gen AI directly without any tip ----  ---- ##### (ELEMENTARY LEVEL) / [Reference 2020](https://arxiv.org/abs/2001.08361)) # Few-shot ### With some examples !!! ----  --- ##### (JUNIOR LEVEL) / [Reference 2022](https://arxiv.org/abs/2201.11903)) ## Chain-of-Thought (CoT) ### Teach GPT thinking. ----  #### Magic words : Step by step. ----  #### (The latest version of GPT3.5 can do thinking now.) #### Thinking = "Do the things step by step" ----  #### CoT with few-shot : make GPT follow your rules. #### (or Self-Consistency) --- ##### (SENIOR / [Reference 202305](https://github.com/dave1010/tree-of-thought-prompting)) ## Tree of Thoughts (ToT) #### Make GPT reviews/rates itself. ---- ##### The very first prompt version of our translation project : ```text Please rate the result and give me a full score of 100 when the generation is completed. You need to check the quality of the translation results like your customers in {Certain Area}. If the point is less than 99, re-generate again until it's bigger than 99. ``` #### (Re-generating if the score is low.) ---- #### Magic words : ```text Imagine three different experts are answering this question. All experts will write down 1 step of their thinking, then share it with the group. Then all experts will go on to the next step, etc. If any expert realizes they're wrong at any point then they leave. The question is ... ``` ----  ----  ##### They do the discussion !!! ----  ---- ##### (SENIOR / [Reference 202307](https://github.com/holarissun/PanelGPT)) ## PanelGPT #### Discussion like a pro  ---- ```text 3 experts are discussing the question with a panel discussion, trying to solve it step by step, and make sure the result is correct and avoid penalty: ``` ---  --- Reference : https://www.promptingguide.ai/