dataset
active
dataset:truthfulqa-benchmark

TruthfulQA Benchmark

817-question benchmark of adversarially constructed questions used to test whether deception features generalize to factual accuracy beyond consciousness self-report

Neighborhood — ranked by edge-count