You are given a short description of the dataset together with examples from the dataset (separated by "--"):

[[ dataset description ]]
{dataset_description}

[[ example questions ]]
{example_questions}


--


Task:
We aim to reformat this dataset so that each response follows a chain-of-thought reasoning style. Your goal is to produce a template that structures the response as follows:
	1.	Task Introduction: Start by introducing the core task given by the dataset description.
	2.	Response Structure: List the logical steps or stages required to solve the type of task given by the dataset description. Each step should be a single, clear sentence. 
	The steps should encourage explicit, structured reasoning (chain-of-thought). 
	3.	General Instructions: If applicable, add any overarching instructions that should be applied to all steps or to the overall reasoning process.

Guidelines:
	•	Focus on clarity, structure, and brevity.
	•	The template you output should only include the task introduction, the list of steps, and any general instructions. Do not include any extra explanations or comments.
	•	If provided with examples from the dataset, you can use them to understand how the questions look like to build the template.