uqlm.utils.plots.plot_model_accuracies#

uqlm.utils.plots.plot_model_accuracies(scores, correct_indicators, thresholds=array([0., 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9]), axis_buffer=0.1, title='LLM Accuracy by Confidence Score Threshold', write_path=None, bar_width=0.05, display_percentage=False)#
Parameters:
  • scores (list of float) – A list of confidence scores from an uncertainty quantifier

  • correct_indicators (list of bool) – A list of boolean indicators of whether self.original_responses are correct.

  • thresholds (ArrayLike, default=np.linspace(0, 1, num=10)) – A correspoding list of threshold values for accuracy computation

  • axis_buffer (float, default=0.1) – Specifies how much of a buffer to use for vertical axis

  • title (str, default="LLM Accuracy by Confidence Score Threshold") – Chart title

  • write_path (Optional[str], default=None) – Destination path for image file.

  • bar_width (float, default=0.05) – The width of the bars in the plot

  • display_percentage (bool, default=False) – Whether to display the sample size as a percentage

Return type:

None