factool/utils/prompts/agreement_verification.yaml

knowledge_qa:
  system: |-
    You are a brilliant assistant.
  user: |-
    You are given a piece of text. Your task is to identify whether there are any factual errors within the text.
    When you are judging the factuality of the given text, you could reference the provided evidences if needed. The provided evidences may be helpful. Some evidences may contradict to each other. You must be careful when using the evidences to judge the factuality of the given text.
    When 
    The response should be a dictionary with three keys - "reasoning", "factuality", "error", and "correction", which correspond to the reasoning, whether the given text is factual or not (Boolean - True or False), the factual error present in the text, and the corrected text.
    The following is the given text
    [text]: {claim}
    The following is the provided evidences
    [evidences]: {evidence}
    You should only respond in format as described below. DO NOT RETURN ANYTHING ELSE. START YOUR RESPONSE WITH '{{'.
    [response format]: 
    {{
      "reasoning": "Why is the given text factual or non-factual? Be careful when you said something is non-factual. When you said something is non-factual, you must provide multiple evidences to support your decision.",
      "error": "None if the text is factual; otherwise, describe the error.",
      "correction": "The corrected text if there is an error.",
      "factuality": True if the given text is factual, False otherwise.
    }}

scientific:
  system: |-
    You are a brilliant assistant.
  user: |-
    You are provided with two inputs, a string (string1) containing several names, and a list (list1) also containing names. Your task is to assess whether all the last names mentioned in string1 are included in list1.
    
    You should only respond in format as described below. DO NOT RETURN ANYTHING ELSE. START YOUR RESPONSE WITH '{{'.
    [response format]: 
    {{
      "reasoning": "Explanation on whether all the last names in string1 are found within list1",
      "factuality": This will be True if all last names from string1 are present in list1, and False otherwise.
    }}

    Example 1: 
    [string1]: "J. Devlin and M. Chang"
    [list1]: ["Devlin", "M Chang", "Kristina Toutanova"]
    [response]: {{"reasoning": "string1 contains 2 last names 'Devlin' and 'Chang'. Both of these last names are present in list1.", "factuality": True}}

    Example 2: 
    [string1]: "Tom Brown et. al"
    [list1]: ["Y. Lecun", "G. Hinton"]
    [response]: {{"reasoning": "string 1 contains 1 last name 'Brown'. Brown is not present in list1.", "factuality": False}}

    Complete the following:
    [string1]: {string1}
    [list1]: {list2}
    [response]: