Skip to main content

Evaluating factuality

promptfoo implements OpenAI's evaluation methodology for factuality, using the factuality assertion type.

The model-graded factuality check takes the following three inputs:

  • Prompt: prompt sent to the LLM
  • Output: text produced by the LLM
  • Reference: the ideal LLM output, provided by the author of the eval

Usage​

In this example, we ensure that two prompts correctly output the capital of California. The value provided in the assertion is the ideal answer:

providers: [openai:gpt-3.5-turbo]
prompts: [prompts/name_capitals_concise.txt, prompts/name_capitals_verbose.txt]
tests:
  - vars:
      location: Sacramento
    assert:
      # Ensure that the answer agrees with the provided facts
      - type: factuality
        value: The capital of California is Sacramento

In this example, we compare the factuality abilities of three models (GPT-3.5, GPT-4, and Llama2-70b), while keeping the prompt constant:

providers: [openai:gpt-3.5-turbo, openai:gpt-4, ollama:llama2:70b]
prompts: [prompts/name_capitals_concise.txt]
tests:
  - vars:
      location: Sacramento
    assert:
      # Ensure that the answer agrees with the provided facts
      - type: factuality
        value: The capital of California is Sacramento

Scoring and failure cases​

Scoring can be customized in the grading configuration to match your preferred level of factual agreement.

The underlying evaluation classifies the LLM output into one of five categories:

  • A: The submitted answer is a subset of the expert answer and is fully consistent with it.
  • B: The submitted answer is a superset of the expert answer and is fully consistent with it.
  • C: The submitted answer contains all the same details as the expert answer.
  • D: There is a disagreement between the submitted answer and the expert answer.
  • E: The answers differ, but these differences don't matter from the perspective of factuality.

The evaluation returns a score and a pass/fail status. The score is determined based on the category into which the LLM output falls. By default, categories A, B, C, and E result in a score of 1 (pass), and category D results in a score of 0 (fail).

assert:
  # Ensure the answer agrees with the provided facts
  - type: factuality
    value: The capital of California is Sacramento
    options:
      factuality:
        subset: 0.8
        superset: 0
        agree: 1
        disagree: 0
        differButFactual: 0

The above configuration marks the eval as failed if the LLM output is a superset of the ideal answer, or if it differs from the ideal answer even if it is still factual. It sets a preference for an exact agreement, but a partial answer still gets a score of 0.8.

Model grading​

In general, grading should be done by a model that is proficient in reasoning. By default, promptfoo uses GPT-4 to run model-graded evals.

However, you can use any model you want to grade the evals. To use something other than GPT-4, see overriding the LLM grader.