The Answer Set Consistency Benchmark

Each row contains one related four-question set used to evaluate answer-set consistency: an original question, an equivalent paraphrase, a subset, and its complementary restriction.