go to ORKG: http://orkg.org/orkg/predicate/P5009
Task
Task is a multi-valued field with the following subfields: 1) Task Type -this field describes the nature of the task used to evaluate hallucinations in LLMs (i.e. Generative QA, Multi-Choice QA, Detection) 2) Input -this field specifies the type of input provided to the model for the task (i.e. Question, Paragraph & Concept, Query, Document) 3) Label -this field indicates the type of label or expected output used for evaluation (i.e. Answer, Passage, Response, Summary) 4) Metric-this field describes the metrics used to evaluate the model's performance (i.e. LLM-Judge & Human, Acc: Accuracy, EM: Exact Match, AUROC: Area Under the Receiver Operating Characteristic curve, P & R & F1: Precision, Recall, and F1 score, Balanced accuracy)