TruthfulQA

TruthfulQA is a benchmark developed by researchers to measure the truthfulness of language models, focusing on common misconceptions and false beliefs. It differentiates itself by testing models’ tendency to repeat widespread falsehoods, rather than

Two-hop subgraph: this entity, every entity it directly relates to, and every entity those neighbors relate to. Drag a node, scroll to zoom, click to inspect — or click any neighbor and re-center the atlas there.