OPEN AI

Two types of large language models (LLMs)

- Base LLM – predicts next word, based on text training data

- Instruction Tuned LLM – tries to follow instructions. Fine-tune on instructions and good
attempts at following those instructions
o Uses LRHF (Reinforcement Learning with Human Feedback)
o Helpful, honest and harmless

GUIDELINES FOR PROMPTING

Tactics:

- Write clear and specific instructions

o Tactic 1 – use delimiters to clearly indicate distinct parts of the output
o Tactic 2 – ask for structured output

o Tactic 3 – check whether conditions are satisfied

 Check assumptions required to do the task
 Ilustración 1Text with instructions Ilustración 2Text without instructions provided

o Tactic 4 – Few-shot prompting

 Give successful examples of completing tasks, then ask model to
perform the task

 .
 .
 .
- Give the model time to think
o Tactic 1 – Specify the steps to compete a task (Step1:… Step2:… StepN:…)
 .
o Tactic 2 – Instruct the model to work out it’s own solution before rushing to a
conclusion